CuriosityStream Inc - Ordinary Shares- Class A Stock

Greetings, and welcome to the CuriosityStream Second Quarter 2026 Financial Results. [Operator Instructions] As a reminder, this comment is being recorded. It is now my pleasure to introduce Vanessa Gillon, Senior Vice President of Operations. Please go ahead.

Thank you, and welcome to CuriosityStream's discussion of its second quarter 2026 financial results. Leading the discussion today are Clinton Stinchcomb, CuriosityStream's Chief Executive Officer; and Brady Hayden, CuriosityStream's Chief Financial Officer. Following management's prepared remarks, we will take questions from the analyst community. But first, I'll review the safe harbor statement. .

During this call, we may make statements related to our business that are forward-looking statements under the federal securities laws. These statements are not guarantees of future performance, but rather are subject to a variety of risks, uncertainties and assumptions. Our actual results could differ materially from expectations reflected in any forward-looking statements.

Please be aware that any forward-looking statements reflect management's current views only and the company undertakes no obligation to revise or update these statements nor make additional forward-looking statements in the future.

For a discussion of the material risks and other important factors that could affect our actual results, please refer to our SEC filings available on the SEC website and on our Investor Relations website as well as the risks and other important factors discussed in today's press release. Additional information will also be set forth in our quarterly report on Form 10-Q for the quarter ended June 30, 2026 when filed.

In addition, reference will be made to non-GAAP financial measures. A reconciliation of these non-GAAP measures to comparable GAAP measures can be found on our website at investors.curiositystream.com. Unless otherwise stated, all comparisons will be against our results for the comparable 2025 period.

Now I'll turn the call over to Clint.

Thank you, Vanessa. Second quarter was a defining quarter for curiosity stream. We delivered the strongest quarterly financial performance in our history, including record operating income, EBITDA, adjusted EBITDA, net income and earnings per share. The results demonstrate what Curiosity can produce when we combine the value of our differentiated content and data assets with our disciplined operating model.

In the quarter, high-value licensing revenue, reliable subscription revenue, efficient spending, lower year-over-year operating expenses and a focused cost structure came together to create substantial operating leverage.

Revenue was $23.2 million in the second quarter, up 22% year-over-year. Licensing revenue was $14.1 million, up 48% from the prior-year quarter. Licensing represented the largest component of our revenue growth in the quarter and highlights the strategic value of the Curiosity corpus and the multiple ways in which we can monetize it.

Today, we're able to engage with leading global media and technology companies through licensing supported by three distinct and durable pillars. First, we licensed premium factual video to broadcast, Paytv, streaming, cable, satellite, wireless and other distribution partners.

Second, we license highly structured custom and off-the-shelf video and audio data sets to technology companies for AI training.

Third, we offer private code corpus of more than 880 billion tokens for licensing to frontier model developers and coding agent providers for AI training and reinforcement learning as well as to enterprises seeking to fine-tune models after pre-training and general training on closed and open source large language models.

While we believe our private code database offering is the largest available in the world, simply put, the onshore volume, we offer unique software engineering environments, containing code, history, decisions, failures, and verifiable outcomes that can improve coding agents through trading, [ RL ] and evaluation.

We believe these three distinct sources of licensing intellectual property reduced medium-term licensing risk and create significant long-term upside. They allow us to participate in several large expanding markets while serving customers with different use cases, buying cycles and commercial objectives.

In response to partner demand, we have also now productized the significant portion of our video library, specifically for AI training. We believe this productization will reduce friction in the licensing process, make it easier for prospective customers to identify and evaluate the data sets they need and ultimately shorten sales cycles.

We currently offer 17 off-the-shelf video data set products. These include extensive premium collections covering scripted entertainment, professional and collegiate sports, animation, wildlife, science, automotive and instructional content as well as highly structured data sets and cliffs built around high dynamic range video, character tracking, synchronized multi-camera footage, emerging objects and raw foods.

Importantly, we're not simply offering large quantities of video. We're increasingly organizing, structuring and packaging our IP around the specific requirements of sophisticated AI developers. We believe this substantially increases both utility and the value of the underlying content.

Approximately $9 million, subscription revenue was roughly equivalent to the second quarter of 2025. We remain committed to our subscription business into building the long-term value of the Curiosity brand and customer relationships.

At the same time, we continue to manage that business for durable economics rather than pursuing growth at any cost. Our diversified monetization model gives us the ability to be disciplined in customer acquisition spending while we capitalize on high-value licensing and distribution opportunities.

We also made meaningful progress in improving the efficiency of the business. By leveraging AI productivity tools and better aligning our talent base with the highest value priorities, we reduced spending across our primary expense categories. Total operating expenses declined 24% year-over-year. We expect to make further progress in the second half of the year.

This is not simply a cost reduction story. This quarter showcases a more efficient business model in which Curiosity can convert high-value revenue into meaningful profitability while continuing to invest selectively in the content, technology, distribution and commercial capabilities that support long-term value creation.

The resulting profitability was exceptional. Net income was a record $8.9 million, up 1,133% compared with $0.9 million in the prior-year quarter. Second quarter EPS was $0.15. Adjusted EBITDA was a record $11.4 million, up approximately 300%. Margins reflected this operating leverage. Gross margin increased to 73% from 53% in the prior year quarter. Adjusted EBITDA margin was 49% compared with 16% in the prior-year quarter.

Our strategy remains clear. We continue to pursue high-value licensing opportunities that recognize the differentiated value of our extensive [ corporate ]. We'll maintain our focus on operating discipline, including thoughtful marketing investment and rigorous expense management. We will continue to build the long-term value of the Curiosity ecosystem across established and emerging platforms. while simultaneously exploiting existing and new grants of rights that we can monetize.

I want to thank the entire Curiosity team for delivering these results. The quarter was a powerful demonstration of the value of our brand, flexibility in our business and the earnings power of the company made possible by the breadth and depth of our IP.

We are pleased with the momentum, but our focus remains squarely on execution and on our longer-term objective, building Curiosity into a company that informs, inspires and entertains and in so doing generates $100 million or more reliable, recurring and increasingly predictable annualized revenue.

I'll now hand the call over to our CFO, Brady Hayden.

Thank you, Clint, and good afternoon, everyone. Our full Q2 results will be in the 10-Q that we'll file in the next day or 2. Let me quickly hit some of our second quarter highlights.

As Clint said, in Q2, we reported revenue of $23.2 million compared to $19 million a year ago. Likewise, we reported record adjusted EBITDA of $11.4 million. This is also our sixth consecutive quarter of positive adjusted EBITDA. We generated second quarter subscription revenue of $8.9 million, a slight improvement from Q1. Licensing came at $14.1 million, a 48% increase from last year.

Second quarter gross margin was 73%, improving from 53% last year, as we were able to generate significant new revenue in the quarter with only minimal incremental distribution costs. Total operating expenses were down by 24.1% as we continue to see the benefits of our ongoing cost rationalization efforts.

We reported record net income in the second quarter of $8.9 million or $0.15 a share. This compares to $0.8 million of net income in the second quarter of 2021. We believe our balance sheet remains in good shape. In June, we paid our regular $5 million dividend, and we repurchased $600,000 of our shares in the quarter.

We also prepaid $2 million to fully consolidate the ownership of our German business and buyout or JV partners, Spiegel and Authentic of their stakes. This transaction officially closed on July 1 and will be reflected in our Q3 results.

We ended the quarter with total cash and securities of $10.9 million and no outstanding debt. Based on our quarterly dividend of $0.085 per share at yesterday's closing price, CuriosityStream shares provide a dividend yield of about 12%.

Looking at our liquidity outlook for the remainder of 2026, we expect to end the year with a cash and investments balance of $17 million to $22 million. We expect revenue for the second half of the year to be $38 million to $41 million and full year 2026 revenue in the range of $77 million to $82 million. Furthermore, we expect adjusted EBITDA for the second half of the year to be $6 million to $10 million and full year 2026 adjusted EBITDA in the range of $18 million to $22 million.

With that, I'll turn it back over to the operator to begin our Q&A.

[Operator Instructions] question, Frank Lorenzo with Singular Research.

It's Frank, but Frankie is fine. Very nice quarter momentum there for the quarter. You talked about the pipeline. Could you give us a little more detail on the pipeline and how that could potentially drive additional AI training and streaming growth related going forward?

Yes. Thanks for the question, Frank. Yes. As it relates to our licensing pipeline, it's certainly as robust as it's ever been. And what gives us confidence there as I mentioned on the call, we have really three distinct and durable licensing pillars that kind of reduce risk and create significant long-term upside, that again, they allow us to participate in large and expanding markets, while serving customers with different use cases, buying cycles and commercial objectives.

And just to restate what those are, we licensed premium factual video to broadcast Paytv streaming, cable, satellite, wireless and other distribution partners. So that's an ongoing licensing business that we have forever and delivers every quarter.

Second, we licensed a highly structured custom and now off-the-shelf video and audio data sets to technology companies for AI training. What we like about these productized off-the-shelf offerings as we believe that will really accelerate sales cycles. We know it accelerates our operational work.

And then third, we offer a private code corpus of more than 880 billion tokens for licensing to AI developers and coding agent providers for a training, reinforcement learning, and evaluation as well as to enterprises seeking to fine-tune models after pre-training and general training on open and closed store LLMs.

So this scope of sets us up very well, it gives us sort of the largest pipeline that we've ever had. On the licensing side, it can be a little bit chunky from time to time. But the scope and scale of our offering today, I think we believe, will enable us to sort of minimize any dips and certainly optimize semi-transformational upside.

Okay. Also, could you talk a little more about international. You did an acquisition there recently. Could you just talk about the international landscape, where are you going to have the potential? Maybe expound upon the expansion plans there and the overall potential for the business?

Yes. Thanks for asking. So on the subscription side, we participated in a joint venture with Spiegel Corporation and with the German company called Authentic for the last few years, and we consolidated our ownership of that in the second quarter. So a meaningful component of our cash and some of our cash went to that.

What we operate within the German-speaking territories are 2 24/7 Paytv channels that have real distribution. We have several distribution with our fast channels as well over there. And if you look across the world, Germany and German-speaking Europe is our largest non-English-speaking market. And so we like the potential and the firmness of that market for us.

And then the nice thing about our content is it's evergreen, it travels well. And a significant portion of our subscribers are international today, and we believe we'll continue to be from outside the U.S. as we go forward. And we believe that as we continue to roll out new currencies, new billing and payment systems, we'll capitalize even more on the opportunity outside the U.S. for us.

Okay. Just one other quick question. I think Disney and a few others have been talking about maybe expanding their streaming offerings, adding streaming from other services, et cetera. Could you kind of talk about that landscape, if there's potential there and if there's room for that regarding consumer streaming budgets?

Well, I think what they're talking about, and we saw it recently with Peacock and YouTube is bundling. And we're big believers in bundling entertainment services, and that's something that you can expect us to continue to pursue aggressively.

And we've put some nice bundles in place over the last 6, 8 months. And we believe that over time, those will provide really sturdy, reliable subscription revenue that helps to maintain our subscription business and also enables us to spend efficiently there. So a big believer in bundles. We'll continue to pursue that aggressively.

Next question, Laura Martin with Needham & Company.

It's Dan stepping in for Laura here. My question is your second half adjusted EBITDA guidance of, what, $6 million to $10 million, does that represents a step down from the $11.4 million that you guys generated? Does the second half moderation reflect higher customer acquisition marketing, cost reinvestment, higher content costs? Or just are you guys just being conservative regarding assumptions for closing additional licensing transactions?

Thank you for that question, Dan. And I would say the latter. We want to take a conservative approach as it relates to forecasting EBITDA. As you can see, we're sort of on a run rate for the year based on the first 2 quarters of $25 million.

We've been heavily focused this year on getting our EBITDA up and over $20 million. And I think that we'll monitor that as the second half of the year goes on. And if it warrants kind of making changes to that guidance, that's something that we will do. But what we'd like to do is meet and exceed our guidance.

Great. I had a follow-up question, if it's okay, on -- so you guys had almost 73% gross margin in the second quarter. And I'm wondering, how much of the 38 to 41 in second half revenue guidance, is contracted -- is already contracted for by [ HI ] training data? Or how much of that is can you say it's going to be like on new deal expectations in Q3, Q4?

Yes, appreciate that question, Dan. We have a lot in the pipeline right now. It's hard to project with great precision where our licensing revenue will end up. But if you look at the first half numbers and say, okay, they're $38 million, $38.5 million for the first half of the year.

If you just took our sort of run rate subscription revenue and other revenue out, that's another $19 million to $20 million, which gets you kind of in that $58 million, $59 million range. The low end of our guidance of $77 million to $82 million would be kind of another $20 million in licensing revenue for the second half of the year. High end would be $24 million. We exceeded that this quarter. So we have confidence in getting to those levels.

And I talked a little bit about what we've done from a productization standpoint, have 17 distinct video data sets today. And that, we really believe, and we're already starting to see it; will help to kind of accelerate our sales cycle. It certainly has helped to accelerate our operational cycle. And as we to do that, we think that, that will make our range certainly a bit more predictable as we go forward.

Next question, Jason Kreyer with Craig-Hallum.

Maybe I'll pick up where we left off, Clint. You were talking about those 17 new off-the-shelf data products. Can you just talk maybe a little bit more about the reception to those products and how that has influenced the pipeline for the second half of the year?

Yes. Thank you for asking that, Jason. And it's not that we're introducing new content, but we productized a significant portion of our library. Again, we have -- we control rights to well over 3 million hours of audio and video across finished programming, raw footage and a host of other types of content.

And so when we mention these off-the-shelf productized data sets, what we're talking about is a distinct data set of scripted entertainment, as an example. There's a certain number of hours there. All of the metadata is baked into it. And when I say all of the metadata, like much more significant metadata than in any kind of video licensing agreement.

We have a video data set of professional and collegiate sports, as an example, that where you might find like well over 100,000 hours of content and some of which is heavily annotated, data sets around science and technology, around animation and anime, around automotive, around wildlife.

So we've got these really distinct sort of historical categories of premium broadcast video. While at the same time, we've built and organized data sets around collections like emerging subjects and objects. And so this is -- these are like clips built around camera reveals, subjects emerging from forests, water, doorways.

We have high dynamic range video, which is really interesting, particularly for video gen companies because today, the model needs to know -- these models need to know a lot more than like just what does an elephant look like, they need to know like what does an elephant look like at noon, at sunset, in the shadow, against bright sky, partially occluded, moving between light and darkness from different camera positions.

So we have these kind of unique categories as well, character tracking, raw footage. And I think if you look at the scope of IP that we licensed in the last quarter, I don't think you can find another company in history that has license sort of the scope of content that we have.

What I mean by that is traditional video to more than 25 platforms and channels. multi-camera synchronized video for AI training, HDR video for model training, millions of tokens of code for more than 10 sources and languages for training. So we have customized data sets. And we have off-the-shelf data sets.

And that doesn't mean that there's not still a lot of evaluation and back and forth as we do these deals. But as you develop -- as you productize more and more of your library end of your code, definitely helps to simplify and accelerate processes that can take some time.

Does that answer your question?

It does. I mean maybe I'll ask a follow-up here. So last quarter, we had talked about how you're engaging more with maybe a new group of LLM and they want to consume content differently than the existing LLMs you're working with. These -- the productization of these off-the-shelf packages, is that meant for this new group of LLMs? Is this kind of the first step in trying to create a monetizable solution for these LLMs? I'm just looking for an update on how that strategy is progressing.

Yes. And I don't mean to imply that there's a lot of big new frontier developers coming into the marketplace. I mean there's 7 or 8 of those guys, and there's a couple of more that kind of sit on the periphery. But beyond those companies, there are an increasing number of companies who need to license either code or video or specific -- some type of specific video or audio to train their models. So the overall number of companies who are licensing IP, that's expanding.

And yes, absolutely helps, especially for some of the -- it helps across the spectrum, but I think certainly for some smaller companies, who might want to take maybe a slightly smaller bite at the beginning than other large companies, it's really helpful.

Just as a company, we just are trying to simplify everything, trying to simplify everything and trying to just bring a level of velocity to everything. And so that's productization velocity or product velocity. That's prospecting velocity, operational velocity, sales velocity, acquisition velocity, deal velocity, marketing velocity.

If you can't bring the velocity, you'll have a hard time kind of lasting here. And I think that's just kind of true across the media landscape. And we just -- I mean, I personally am just kind of I'm amazed every day at the sort of world that we're living in today. I'm so grateful to be living and operating in this time when there is just extraordinary opportunity in front of us.

And in our case, we have the opportunity to engage with the world's largest companies and many of the most exciting emerging companies.

I'm going to ask one more. Just going back to the numbers, really good gross margin improvement this quarter, really good cost management this quarter? And maybe reconcile that back to Dan's question, just in terms of if we look at the second half guide, if we continue gross margins where they're at and we continue the cost discipline on OpEx; it seems like that would produce a bigger EBITDA number. So just wondering, how we should think about both margins and OpEx as we get into the second half?

Yes. Well, I think you can definitely think about OpEx as decreasing. So a lot of the work that we've done over the last year, you'll see in the second half of the year. So -- whereas this last quarter, I think our OpEx overall expenses were down about 24%. And I think if you look at the second half of the year as compared to the first half of the year, there will be an additional probably 18% to 20% reduction there in our costs.

So we think that there'll be obviously some EBITDA opportunity on the cost side. And then as it relates to the revenue side, there will probably be more rev share agreements in the second half of the year than they were in the first half of the year. And so we've just kind of tried to balance out the EBITDA based on those projections.

Thank you. This does conclude today's teleconference. You may disconnect your lines at this time, and we thank you for your participation.

Good afternoon. My name is Stacy, and I will be your conference operator today. At this time, I would like to welcome everyone to the CuriosityStream First Quarter 2026 Earnings Conference Call.

Please note that today's call is being recorded. [Operator Instructions] I will now turn the call over to Vanessa Gillon, CuriosityStream's Senior Vice President of Operations. Thank you. You may begin.

Thank you, and welcome to CuriosityStream's discussion of its first quarter 2026 financial results. Leading the discussion today are Clint Stinchcomb, CuriosityStream's Chief Executive Officer; and Brady Hayden, CuriosityStream's Chief Financial Officer.

Following management's prepared remarks, we will take questions from the analyst community. But first, I'll review the safe harbor statement. During this call, we may make statements related to our business that are forward-looking statements under the federal securities laws. These statements are not guarantees of future performance, but rather are subject to a variety of risks, uncertainties and assumptions.

Our actual results could differ materially from expectations reflected in any forward-looking statements. Please be aware that any forward-looking statements reflect management's current views only, and the company undertakes no obligation to revise or update these statements nor to make additional forward-looking statements in the future. For a discussion of the material risks and other important factors that could affect our actual results, please refer to our SEC filings available on the SEC website and on our Investor Relations website as well as the risks and other important factors discussed in today's press release.

Additional information will also be set forth in our quarterly report on Form 10-Q for the quarter ended March 31, 2026, when filed. In addition, reference will be made to non-GAAP financial measures.

A reconciliation of these non-GAAP measures to comparable GAAP measures can be found on our website at investors.curiositystream.com. Unless otherwise stated, all comparisons will be against our results for the comparable 2025 period. Now I'll turn the call over to Clint.

Thank you, Vanessa. Financially, we remain focused on building CuriosityStream into a company with $100 million or more reliable, recurring and increasingly predictable annualized revenue. We've done a substantial amount of foundational work to position the company for that objective, and we believe that with continued execution, the path is becoming clearer.

To optimize toward that milestone target, we made several deliberate choices in Q1 that affected near-term quarterly revenue, but in our view, strengthened the company's medium- and long-term revenue opportunity. This is why we guided to the first half of the year as compared to the first quarter.

First, we entered into pilot and framework agreements with certain large-scale partners covering broader and more valuable data sets. This meant prioritizing the structure, scope and expansion potential of the relationship over maximizing upfront revenue recognition in the quarter.

We believe that was the right trade-off. These pilot structures give partners a path to test, validate and scale across a deeper and wider range of CURI assets, which we believe will lead to larger, more durable licensing relationships over the next year and beyond.

Second, we made some modest technology investments that while not required to operate our business in Q1, we believe will enable us to accelerate provisioning and expand and maximize the scope, scale, specificity and profitability of upcoming partnerships.

Third, we developed and organized more licensable IP at considerable scale, most of which is 100% owned. This is important strategically as diversity of data and full ownership of more IP in our corpus strengthens margins, broadens our addressable partner roster, increases revenue potential, reduces reliance on any single transaction and makes the licensing business more predictable over time.

For Q1, revenue was $15.2 million, up slightly year-over-year. Subscription revenue was roughly equivalent to the prior quarter. AI licensing revenue, which we have previously said would be lumpy, was indeed lumpy. And this was in light of the pilot-oriented approach we adopted during the quarter.

Importantly, while the near-term revenue impact may not be immediately obvious, we entered into agreements with a broader roster of partners than we had at this stage last year. We view that as a meaningful indicator of demand and market validation.

The breadth of assets now being discussed and packaged for partners has expanded considerably and includes hundreds of millions of production-grade temporal ground truth tokens for frontier model training and tuning, HDR video, matched raw and finished video, multi-camera video and ego-centric video for physical AI training.

Licensing revenue does not always move in a straight line quarter-to-quarter, especially when we are dealing with larger partners, new partners, broader rights packages and more complex data products. Further, our corpus is built on assets that are scarce, rights-aware, difficult to replicate and increasingly valuable.

We are not talking about a single opportunistic window. We are talking about a monetization model anchored in premium, unscripted and scripted media, enriched structured metadata, flexible rights and growing demand from AI developers and traditional media companies.

CuriosityStream has built a large differentiated content library of rights to over 3 million hours of premium factual content plus sports, plus news, plus general entertainment, animation and film, finished in raw, ego-centric and multi-camera, supported by more than 200 content and data partners and flexible licensing rights. This is not commodity inventory.

It is a scaled, unscrapable, curated corpus that took years of capital, relationships, editorial focus and dense work to assemble. We don't believe it makes sense to discount it for temporary gain. Enduring revenue streams are almost always rooted in assets that are hard to replace and expensive to rebuild.

Looking ahead, Q1 sequential revenue decline was anticipated and we believe temporary. We currently expect 2026 to represent a significant step-up in both revenue and cash flow compared to 2025, with subscription revenue increasing by single-digit percentages and with licensing becoming the larger growth engine as it surpasses subscriptions for the full year.

Several factors support this outlook, the impact of our new pricing and packaging, which is just beginning to roll through our P&L, a solid partner launch pipeline with dominant global distributors, accelerating AI licensing fulfillments, new partner additions, continued expansion of our corpus and the ramp of advertising opportunities.

Traditional media licensing remains healthy and diversified, while AI demand continues to broaden across model refresh cycles, enterprise fine-tuning, multimodal applications, source code, physical AI, video understanding and the need for premium rights-aware structured data.

Q1 was a transition quarter in which we deliberately chose to build for larger, broader and more durable licensing opportunities. We believe those choices position CuriosityStream to generate stronger revenue, higher cash flow and greater shareholder value as we move through 2026 and beyond.

In summary, we believe that we will continue double-digit growth in both revenue and cash flow driven by subscriptions and licensing expansion. We continue to reduce expenses through nonessential eliminations in the embrace of evolving AI-infused productivity tools. While we are raising our quarterly dividend of $0.05 to $0.085, we intend to pay 2026 dividends from cash generated by operations as we did in 2024.

Our balance sheet remains strong with over $23 million in liquidity and no debt, giving us substantial flexibility. I'll now hand the call over to our CFO, Brady Hayden.

Thanks, Clint, and good afternoon, everyone. Our full results will be in the 10-Q that we'll file within the next few hours, but let me hit some of our first quarter highlights. As Clint said, in the first quarter, we reported revenue of $15.2 million, a slight improvement compared to $15.1 million a year ago. Likewise, we reported what is now our fifth quarter of positive adjusted EBITDA, which came in at $0.9 million.

Adjusted free cash flow came in at $1.3 million, which represented our 9th consecutive quarter of positive adjusted free cash flow. We generated first quarter subscription revenue of $8.8 million, roughly equivalent to Q4 results. Licensing came in at $6 million, an increase of 11% from last year.

First quarter gross margin was 56%, improving from 53% last year. While distribution costs were lower during the quarter, we invested in certain technology products that led to an increase for the quarter, but from which we believe we will incur lower fees going forward.

Combined costs for advertising and marketing plus G&A were higher by 27% compared to last year. This increase was driven by a noncash charge for stock-based compensation of $2.2 million or about $0.04 on a per share basis and to a lesser extent, slightly higher advertising costs associated with new customer acquisition investments in the quarter.

We reported a first quarter net loss of $1.3 million or $0.02 a share. This compares to a $0.3 million net income in the first quarter of 2025. While our revenue was up from last year, the net loss was driven primarily by the noncash SBC. And as we said earlier, adjusted free cash flow was $1.3 million in the quarter, representing our ninth consecutive quarter of positive results in this metric.

We believe our balance sheet remains in great shape. In March, we paid our regular $4.9 million dividend while buying back $300,000 of our shares, and we ended the quarter with total cash and securities of $23.4 million and no outstanding debt.

Based on our new quarterly dividend of $0.085 per share at yesterday's closing price, CuriosityStream is generating a dividend yield of over 11%. Looking ahead, we anticipate consolidating our ownership of our German business, buying Spiegel and Authentic out of their stakes sometime in the next few months. Purchase price will be approximately $1.9 million, anticipate the transaction will be accretive to earnings.

Regarding guidance, in response to investor recommendations, we're changing up and expanding our metrics going forward for 2026. For the first half of this year, we expect revenue in the range of $35 million to $41 million and full year 2026 revenue in the range of $75 million to $80 million.

Likewise, we expect adjusted EBITDA for the first half of the year to be $5 million to $7 million and full year 2026 adjusted EBITDA in the range of $16 million to $20 million.

With that, I'll turn it back over to Vanessa to begin our Q&A.

Thank you, Brady. We will now turn to questions from the analyst community, including Patrick Sholl from Barrington Research, Jason Kreyer from Craig-Hallum and Laura Martin from Needham & Company. Starting first with questions regarding our subscription business.

From Pat Sholl, what subscriber trends are you seeing? And is it difficult to retain subscribers given your focus on AI licensing? And then from Jason, what has been the response to your price increase on the streaming service? Clint?

Thank you, Vanessa. Our licensing initiatives have no impact on our subscriber retention. Both licensing and subscription businesses require new content typically. So in that sense, they are synergistic. But we're able to attend to both businesses, both revenue streams with the resources at hand. Our robust licensing business with the hundreds of partners who are part of our corpus only helps our retention efforts as it enables us to deploy much more video on our traditional platforms.

Seeing positive response to the price increase, minimal churn and an increase in lifetime value.

Okay. Next question also from Pat. Any update on subscriber acquisition focus between bundled versus direct?

I appreciate that question. We optimize for overall subscription revenue growth and are largely agnostic as to the source. Pure direct subscribers offer the highest ARPU, channel store subscribers offer a lower CPA and wholesale subscribers, while lower ARPU provide a longer-term recurring revenue stream. We value subscribers in whatever channel we reach them.

Okay. Now on to licensing. Also from Pat, can you provide any further information on how content that is licensed is being valued in more recent renewals?

Thank you, Pat, for that question. As we know, overall CapEx for the 5 largest technology companies will be north of $1 trillion in 2027. About 2% to 5% of that will go to data set training. But in regard to pricing, certain content is valued at a higher unit rate.

Content like scripted entertainment, sports, HDR, selective clip natural history and deeply processed content all command higher unit rates than, for example, unprocessed raw content delivered in bulk. What we notice about value is that orders have become increasingly bespoke, increasingly specific.

The comment I would bring attention to here is that CURI has it all, content across the waterfront of genres, meaning we have documentary video content, of course, but we have scripted television and movies in many languages. We have sports across the board. We have audio, we have egocentric, we have HDR. We have it all for licensing to train models. The harder to find the content, the higher the price tag.

Thank you. And from Jason Kreyer, Craig-Hallum. Can you talk about the breadth of AI licensing, how you view the size of the market, how content is being valued and how the AI pipeline is tracking?

Yes. I appreciate that question, Jason. The breadth of assets now being discussed and packaged for partners has expanded considerably and now includes hundreds of millions of production-grade temporal ground truth tokens for frontier model training and, HDR video, match rod finish video, multi-camera video and egocentric video for physical AI training.

A broad set of video and data and a broader set of partners smooth things out over time. But as I mentioned, licensing revenue does not always move in a straight line quarter-to-quarter, especially when we are dealing with larger partners, broader rights packages and more complex data products.

Okay. A couple of final questions on AI, both from Laura Martin, Needham. Clint, can you speak to the longevity of the AI licensing business? What are you seeing in existing relationships?

Also, in light of the revenue and your leadership position in AI licensing, what is your rationale for staying in the subscription business?

Thank you for that, Laura. Some of our best customers are some of our oldest customers, in part because their orders are broadening. They're broadening out to include the additional data sets I just described. We are fulfilling our 10th delivery next week with one of our early partners.

The need for content for fiscal AI training, largely egocentric and egocentric-adjacent offers a set of another 50 companies who will or are engaging in data set licensing. In regard to our subscription business, it is key to our overall strength because it's reliable, recurring and predictable.

It's already built. The barriers to entry are considerable, meaning others would need to spend substantially in order to achieve what we have in hand. And our overall subscription revenue is growing modestly to mid-single digits this year.

Our subscription business takes nothing from our licensing business and is, in fact, critical to the amassing of our 3 million-plus hour war chest and synergistic insofar as our relationships, both with content partners and the buyers cover all of our businesses, subscription, licensing and advertising.

Clint, Brady, thank you both. This concludes our Q&A, and I will now hand it back to the operator.

This concludes today's teleconference. You may disconnect your lines at this time, and we thank you for your participation.

Greetings, and welcome to the CuriosityStream Fourth Quarter and Year-end 2025 Results Conference Call. [Operator Instructions]

It is now my pleasure to introduce your host, Tia Cudahy, Chief Operating Officer. Thank you. You may begin.

Thank you, and welcome to CuriosityStream's discussion of its fourth quarter and full year 2025 financial results. Leading the discussion today are Clint Stinchcomb, CuriosityStream's Chief Executive Officer; and Brady Hayden, CuriosityStream's Chief Financial Officer. Following management's prepared remarks, we will be happy to take your questions. But first, I'll review the safe harbor statement.

During this call, we may make statements related to our business that are forward-looking statements under the federal securities laws. These statements are not guarantees of future performance, but rather are subject to a variety of risks, uncertainties and assumptions. Our actual results could differ materially from expectations reflected in any forward-looking statements. Please be aware that any forward-looking statements reflect management's current views only, and the company undertakes no obligation to revise or update these statements nor to make additional forward-looking statements in the future. For a discussion of the material risks and other important factors that could affect our actual results, please refer to our SEC filings available on the SEC website and on our Investor Relations website as well as the risks and factors discussed in today's press release. Additional information will also be set forth in our annual report on Form 10-K for the fiscal year ended December 31, 2025, when filed.

In addition, reference will be made to non-GAAP financial measures. A reconciliation of these non-GAAP measures to comparable GAAP measures can be found on our website at investors.curiositystream.com. Unless otherwise stated, all comparisons will be against our results for the comparable 2024 period.

Now I'll turn the call over to Clint.

Thank you, Tia, and good evening, everyone.

CuriosityStream was built on one timeless idea. Curiosity changes the world, that every breakthrough begins with a question. 1,000 years ago, Leif Ericsson sailed West into the unknown and discovered a new world. Nearly a millennium later, Neil Armstrong stepped on to the lunar surface, carrying the same enduring message across time. Discovery belongs to the bold and curiosity is our compass. Promotion waves to moon dust, that spirit propels us forward today.

In that same spirit of bold exploration, we delivered strong full year 2025 results. Revenue grew 40% to $71.7 million from $51.1 million in '24, while adjusted free cash flow increased 46%, $13.9 million from $9.5 million in '24. Q4 revenue rose 36% year-over-year to $19.2 million from $14.1 million and adjusted free cash flow climbed 33% to $4.3 million. These gains reflect the strength of our complementary revenue pillars, licensing, driven by high-volume and heavily structured video fulfillments for AI model training, subscription sturdiness through operational execution and new partnerships, amplified by cost discipline that expanded gross margins to 60% in Q4 from 52% a year ago and reduced nondiscretionary G&A expenses by 33% year-over-year.

In 2026, we believe our annual licensing revenue will exceed our overall subscription revenue. We believe we will grow our subscription revenue by low to mid-single-digit percentages because of three key drivers: new pricing, which we began rolling out March 1, new wholesale and retail partnerships, and organic growth from existing partnerships. The recurring, reliable and predictable revenue from our subscription services cements our foundation.

Why do we believe we will see licensing revenue eclipse subscription revenue in 2026? Why do we believe licensing will be robust and durable for the foreseeable future? What is the impact to top line, bottom line and margin expansion? Well, we've covered some of this before, many investors, analysts and commercial partners tell us it bears repeating.

CuriosityStream's licensing business is durable because it's built on assets that are durable, that are scarce, rights aware, difficult to replicate and increasingly valuable across multiple end markets. We're not talking about a single opportunistic window. We're talking about a monetization model anchored in premium unscripted and scripted media, enriched structured metadata, flexible rights and growing demand from AI developers and traditional media companies. CuriosityStream has built a large differentiated content library of rights to nearly 3 million hours of premium factual content plus sports, plus news, plus general entertainment, animation and film, finished and raw supported by more than 200 content and data partners and flexible licensing rights.

This is not commodity inventory. It is scaled, unscrapable, curated, a corpus that took years of capital relationships, editorial focus and dense work to assemble. Enduring revenue streams are almost always rooted in assets that are hard to replace and expensive to rebuild. Demand is broadening, not narrowing. Beyond repeat business from existing customers, we expect our overall roster of partners to more than double in 2026 and potentially increase 5 to 6x in 2027 as the fine-tuning of open source and certain proprietary models opens opportunities for thousands of companies.

Historically, licensing meant selling finished programs or package rights to broadcasters, streamers and PayTV partners. That business remains alive and healthy. And in 2025, we announced new licensing agreements with linear broadcasters, educational platforms, digital-first outlets, global streaming services and, of course, next-generation AI training developers. This diversification makes licensing more durable and cycle resilient. Traditional media licensing is healthy and not going away, but AI licensing is accelerating much faster and driving the bulk of our growth here.

Over the next 5 years, AI model development, model refresh cycles, geographic expansion, enterprise fine-tuning, education applications, agentic systems and multimodal search should all support continued appetite for premium licensed corporate.

For AI license partners as their model sophistication grows, so does the need for more video inputs. Developers require large volumes of high-integrity rights aware training inputs. Premium broadcast video, clean audio, scripts, captions, study guides, metadata and derivative assets have utility well beyond entertainment viewing. They help train, tune, evaluate, ground and improve multimodal systems. The more advanced models become the more they need high-quality structured, legally licensable data rather than undifferentiated scraped material.

So key to note that rights cleared, structured content will become more valuable over time, not less. There's plenty of media on the open web, but much of it is noisy, duplicative, poorly labeled, low-quality or legally ambiguous. By contrast, CuriosityStream's corpus is assembled, curated and increasingly productized for commercial use cases. The premium quality of our video also helps us stand out as we have video captured with top-tier equipment like red cameras, HDR formats and Blackmagic workflows, delivering cinematic excellence with real-world visual depth. This means sharp, high-resolution footage that captures subtle details from the textures of ancient ruins in history to the fluid motions and wildlife sequences.

For AI training, this translates to superior data for tasks like object recognition, scene understanding and generative video. Said plainly, we generate competitive escape velocity through our expanded data structuring and metadata capabilities that are designed to meet partner volume requirements and bespoke specifications. We're not merely selling files, we're not merely selling clips, we are selling usable data sets. That distinction is critical. In AI, a rights cleared file has value. Our rights cleared file with strong metadata, taxonomy, providence, segmentation and packaging has much more value. That creates pricing power and maintenance.

Further, our licensing model benefits from operating leverage and the fact that the standard industry licensing practice in the AI space is one of non-exclusivity. I cannot emphasize enough the value of this dynamic. As our critical mass corpus is now assembled and the infrastructure is largely in place, each new partnership carries attractive incremental economics as our hard cost to create a license in content are largely de minimis. We have, and we will continue to increase our volume through a rev share construct that minimizes cost and risk. We can now monetize the same video multiple times in multiple forms across multiple geographies and buyer classes.

Of course, durability does not mean inevitability. We have to execute. We have to move the ball forward every day. We need to continue acquiring, negotiating sufficient scopes of rights enriching made the data segmenting our corpus intelligently, protecting quality and packaging assets in ways that map directly to buyer workflows. We need to stay disciplined on pricing and avoid treating a library like an undifferentiated commodity supply.

We also need to manage legal and policy development thoughtfully. But all of these are execution challenges. These are not reasons to doubt the model. In fact, a market that increasingly values providence, trust and rights discipline should favor CuriosityStream, not hurt it.

Our view is informed. Our view is straightforward. CuriosityStream's licensing of video, audio, images, scripts and related data products is durable because it rests on scarce assets, diversified demand, strong reuse economics and a market shift toward high-quality licensable content. It can continue to grow significantly because we are still early in the monetization curve. It will be lumpy over 3- and 6-month tranches. But as Warren Buffett often said, we'd rather have a lumpy 15% than a smooth 12%.

Traditional licensing is meaningful. AI licensing is scaling rapidly. And the strategic value of curated rights to aware, metadata-rich premium media compounds over time. This is why we believe licensing will remain a critical and durable growth engine for the long-term foreseeable future.

In summary, we believe that we will continue double-digit growth in both revenue and cash flow driven by subscriptions and licensing expansion. We intend to pay 2026 dividends from cash generated by operations as we did in 2024. Our balance sheet remains strong with over $27 million in liquidity and no debt, which we believe gives us substantial flexibility.

I'll now hand the call over to our CFO, Brady Hayden, who I'm sure will emphasize that among other attributes, at today's share price, we're a growth company that also offers a dividend yield of 10%.

Thank you, Clint, and good evening, everyone. Our full financial results are presented in the back of the press release that we just issued a few minutes ago as well as the 10-K that we'll file in the next few days. But let me quickly go through some of the results that we want to highlight for the fourth quarter as well as full year 2025.

In the fourth quarter, we reported revenue of $19.2 million at the high end of our guidance and a 36% increase compared to $14.1 million a year ago. For the full year, revenue was $71.7 million, a 40% increase from last year. Likewise, we reported another quarter of positive adjusted EBITDA, which came in at $1.1 million. This was an improvement of $3.1 million from a year ago, and also our fourth sequential quarter of positive adjusted EBITDA. For the full year, adjusted EBITDA was $8.2 million, a $14.3 million improvement from 2024.

Adjusted free cash flow exceeded our guidance in the fourth quarter at $4.3 million, which is also our eighth consecutive quarter of positive operating cash. For the full year, adjusted free cash flow was $13.9 million, a 46% increase from $9.5 million in 2024. Licensing revenue was $9.8 million in the fourth quarter, an increase of $6.1 million from last year, while subscription revenue came in at $9.1 million. For the full year, subscriptions were $37 million, while licensing came in at $33.2 million. This was an increase of over $25 million in 2024 and driven by continued growth in AI training fulfillments.

Fourth quarter and full year gross margins were 60% and 57%, respectively, each of these improving from last year. Within cost of revenue, storage and delivery costs increased during the year in light of the high volume of video we put into AI licensing agreements. For the full year, combined costs for advertising and marketing plus G&A were higher by 24% compared to last year. Although this increase was the result of noncash charges for stock-based compensation of $14.4 million or about $0.24 on a per share basis.

G&A also included an adjustment to payroll costs for incentive compensation as well as a number of onetime expenses associated with our August secondary stock offering. Were it not for the noncash SBC, the incentive comp adjustment and the common stock sale, G&A would have declined by over $1 million in 2025.

For the full year, net loss was $6.4 million compared to a net loss of $12.9 million in 2024, representing an improvement of over 50% in net loss. While our revenue was up materially from last year, the 2025 net loss was driven by the onetime charges, incentive comp adjustment and noncash SBC. Were it not for these specific charges, we would have posted positive earnings for the year.

And as we said earlier, adjusted EBITDA was $1.1 million in the fourth quarter compared to a loss of $1.9 million a year ago. And for the full year, adjusted EBITDA was $8.2 million compared to an adjusted EBITDA loss of $6 million in 2024.

For the full year, adjusted free cash flow was $13.9 million, a 46% increase from $9.5 million in 2024. This totals well over $20 million in operating cash that we've generated over the last 2 years. On October 14, $6.7 million of our warrants expired unexercised. All these warrants have been trading well out of the money for some time, this expiration of all of the company's outstanding warrants reduces potential dilution and should eliminate any lingering share overhang associated with these instruments.

In December, we paid $4.7 million for our fourth quarter dividend, including our $0.10 special dividend paid in June, this brings our total dividends paid to $22 million for all of 2025. We ended the year with total cash and securities of $27.3 million and no outstanding debt, and we believe our balance sheet remains in great shape.

Based on yesterday's share price, CuriosityStream is generating an adjusted free cash flow yield of over 8% and a current dividend yield of over 10%. Given where our shares have recently been trading, we just announced that our Board has increased our share repurchase authorization to $6 million we plan to selectively resume our repurchase activity in the coming weeks and months.

Moving to guidance. For the first half of 2026, we expect revenue in the range of $38 million to $42 million and adjusted free cash flow in the range of $6 million to $9 million. For the full year, we continue to believe we'll achieve double-digit growth in both revenue and cash flow in 2026 and that a full year of positive GAAP earnings is achievable.

With that, we can hand it back to the operator and open the call to questions.

[Operator Instructions]

We're ready to take questions. Patrick Sholl from Barrington, you're up first. Go ahead.

[Technical Difficulty]

This is the CuriosityStream year-end 2025 conference call, experiencing some technical difficulties. Thank you for continuing to hold. We will be with you as soon as we can. Again, appreciate you continuing to hold.

Thank you again for your patience. This is the CuriosityStream year-end 2025 earnings call. We are still having technical difficulty. If you are in the question queue right now, would you please e-mail your questions and reply to the e-mail that you are about to receive.

And we will take questions over e-mail shortly. Thank you so much, and thank you for continuing to hold.

Thank you for holding. This is the CuriosityStream 2025 year-end earnings report. Our first question today comes from Dan Medina from Needham. The first question is, could you please update us on whether LLM licensors are renewing their deals with you and how the nature of second contracts is different from the earliest LLM contracts you licensed.

Thank you, Dan, for that question. Really appreciate it. The answer is yes. Virtually everyone has renewed or will renew. And the beauty of the second agreement is it's always easier because you have the paper in place. Same thing with the third agreement, same thing with the fourth fulfillment. So without a doubt, we're seeing repeat business. At the same time, we're seeing a lot of new potential partners express interest and express very high volume and specific requirements that were working aggressively to fulfill right now.

A second question from Dan. Any change in the pace of adding other companies' libraries to your ability to license ours to the LLM.

Yes. We like the Golden Gate Bridge there, Dan. We are constantly in acquisition mode. We've built in the mass, I think, an extraordinary library. We we've been told, just told this week by the most valuable by market cap companies in the world that we have the best video corpus for AI training. So we have video in place. We have paper in place with the world's biggest companies. We've enhanced our human talent. We've done the necessary things to ensure the steadiness of our subscription services and feel really, really good about the year, Dan.

And one final from Dan. Can you give us some cases of how LLMs are using the information you license to them in the market to make money, tools and apps.

I think that's a great question, Dan. And I think that if you look at sort of the evolution of -- what our technology partners are looking for and are working toward, I think, if you start sort of 2020 with large language models. There was a lot of text that started there and that was designed to help teach the models to read, to help create document summarizers, knowledge, Q&A, support bots, active coating copilots. I think we to transition up to scale to kind of multimodal AI, which is text, which is images, which is audio, which is video and that led to sort of video summarization, camera assistance, text image, text to video sort of in the agentic AI.

Obviously, it's part of the spectrum now. And that's where systems that plan use tools and act autonomously. And the sort of use cases there are research agents, travel booking assistants, code agents, data ops agents, CRM bots, there's almost an infinite number of use cases. And then I think, certainly, an exciting stage that we're early stages of right now, it's physical AI, where the content is being used to kind of embed AI into robots into cars, into drones, into devices.

And similarly, I think, an infinite number of use cases here, warehouse robots, self-driving cars, home robots, delivery drones, factory arms, all kinds of things. So extraordinarily exciting, difficult to stay up with all of the use cases. But the good news is we have such a variety such a strong scope of video and data that we're able to fill a large scope and scale of requirements.

Thank you, Clint. Our next question is from Jason Kreyer from Craig-Hallum. He says, Clint, you called out 2026 as the greatest year in company history. Can you unpack that from a metrics standpoint, perhaps with some more clarity on your goals for the base streaming business and then the licensing opportunity.

Yes, thank you for that question. Jason, so we made a lot of progress in 2025. We talked about the increases in cash flow and top line revenue in the size of our library and the quality of our library. And so as it relates -- as our subscription business is concerned, and that includes wholesale and retail subscriptions. We're confident that we're going to grow that at low to mid-single digits, and we're really confident in that because we have new partnerships coming on every month. And with channel stores around the world for Curiosity, for Curiosity, and even for Catholic stream.

We have new wholesale relationships that are that are rolling out over the next several months and even now. And we took a price increase March 1, and that's going to take a while to roll through our financials. But with those three things and with the marketing money that we're spending, like we're very confident that we'll grow our subscription business in the low to mid-single digits. And so based on, I think, what Brady shared, that's off a base of $36 million, $37 million a year.

On the subscription side, we're confident -- I'm sorry, on the licensing side, we're confident that we're going to eclipse our subscription revenue because of the work that we've done to date. We are experiencing and anticipating a lot of repeat business or business from existing partners and customers. And at the same time, I think that our new Chief Commercial Officer, John Vilade, has brought extraordinary amount of kind of velocity to our efforts right now. And so in working with our key people here, our ops team, we're going sort of way beyond the obvious top 6, 8 companies that are in the space and anticipate expanding our roster really significantly this year.

Now again, that's going to be -- that will be choppy, but the opportunities are big. I think one of the most exciting -- I'm glad that I lived or to kind of work through this period of time because we've got the goods. We have really unique advantages. We need to execute, but I've never in my career have been so close to so many big opportunities at the same time.

And thank you, Clint. Some questions now from Dave Marsh from Singular Research. On the subscription front, how many new platforms are you expecting to launch during FY '26? And how many new countries do you think you could launch with existing partners?

Yes. Great question, Dave. Thank you. Well, I think just this -- in just this year alone, we've already launched with Apple in Canada as one of many examples. And we anticipate that over the course of this year, probably 12 to 20 new platforms. Some of these are -- they're not all created equal. Some are larger than others, some delivered more opportunity than others, but certainly 12 to 20 over the course of this year. And the beauty of all of that is the partners that we're working with are good at growing subscribers. So we feel really confident about our ability to grow that side of the business, and it's dirty.

And the second question from Dave Marsh from Singular Research. If I heard you correctly, it sounded like SG&A would have been down $1 million year-over-year without the nonrecurring charges. So with mid-$20 million be a good expected run rate for fiscal year '26.

Can you take that?

Yes, I can take that. And good question, Dave. We Yes. We don't provide guidance on the expense side. But I think those are fair numbers obviously, with stock-based comp. That's a little bit of a wildcard because of the way we and we got -- I got into this in the last call. But the way we award our grants and the way the accounting treatment is applied to those, it can be somewhat difficult to predict. But I think if you take out stock-based comp, we're actually looking at G&A other than SBC below 20%. So I think your range is certainly fair.

And one final question from Dave March. Any M&A opportunities you might consider?

Thanks for that question, Dave. We'll always do what's in the best interest of our shareholders. I think that the M&A world environment will be exciting this year. We'll be right. I think if you look at some of the deals that have been done most recently with the big companies, those are a lot more around synergies. But we believe that if we continue to execute, continue to post good increases continue to show the value of our subscription business of our licensing business that we'll have the opportunity to consider whatever combinations are in the best interest of our shareholders.

Clint, our next questions are from Pat Sholl, Barrington Research. Could you provide any additional color on the market for content to license for AI training and how your partnership with Versos video training linear [indiscernible].

Yes. So Versos is it's a really good company. They are a technology partner of ours. We've worked with them for a long time. I mean they help us to organize our content for the most part, help to clip our content and just -- they help us manage an increasingly sort of large volume of content as we are organizing fulfillments there. Now we did a lot of licensing agreements before we started working with Versos, but I think they're helping us to -- help with by handling some of the work on the organization side, helping us to do even more.

As far as the content that we offer today. I think a lot of people rightly think of CuriosityStream as a company focused in the factual media space. And certainly, we are. And certainly, we have a variety of content there. I mean we have a corpus today that is a collection of content from not just ourselves but from over 200 partners. And so in addition to the full range of factual content, prime high -- historical prime, SBNs, travel food culture home. We also have a good corpus of scripted content, which is really hard to acquire for a variety of reasons.

Dramas, comedies Western's action films, adventure films, mystery, family faith film, et cetera. And we also have a broad collection of sports, American football, soccer, surfing, tennis, basketball, billiards, boxing, drifting, a lots of combat sports. So we have a full corpus there. So that's something that gives us a unique advantage and enables us to just engage with virtually everybody on the planet who has video licensing needs for training and other purposes.

And the second question from Pat Sholl, Barrington Research. With the price increase implemented March 1, what is the timing of it being fully implemented and expectations on churn?

Yes, it will take a year to fully implement just because we have so many people on annual subscriptions. I think what you'll see in the first month is probably 3% to 4% of all of our customers, 5% maybe who are -- who become part of that. And so that will roll out over time with our pure direct customers, obviously, won't roll out fully until everyone has renewed their agreement.

On the partner side, most of those subscribers are monthly. And there -- it takes some of them a little bit longer to roll out the pricing increase. But we anticipate that over the next handful of months, everybody will. So we'll get significant benefit this year, and we'll continue to get benefit through February of next year.

And the final question from Pat Sholl. Any additional commentary on the cadence of guidance and expectations on the full year?

Yes. I'll speak to that for a minute, and then I'll hand over to Brady for his point of view as well. We guided to the half year because the many of the partnerships that we're working on are just -- are large and have the potential to be very large. And in the a little bit lumpy. I mean the benefit to working with the biggest companies in the world is you know you're going to get paid. You're not chasing people to get paid. However, sometimes the payment schedules can be a little different than certain other companies.

So the cash revenue can be a little bit lumpy as it relates to this in light of these big licensing opportunities. And so we're extremely confident in the year that we're going to have this year without giving specific year-end guidance, like we said, our intent is to is to pay our dividend from cash from operations. And our belief is that our licensing revenue will exceed our subscription revenue. So feel good about where we're going to end up. We've said double-digit increases in both cash flow and top line revenue, and that's what we're working toward every day and confident that we'll achieve.

Yes. The only thing I'll add is the revenue cycle for these deals, and we've talked about this before, but it's generally between 4 and 6 months, we're delivering content, we're then recognizing the revenue. We go through an acceptance process. We will -- we're not issuing our POs until we're actually getting paid under most of the contracts that we're doing. So the entire cycle can last as long as 6 months, and it's just become very difficult for us to predict with much precision exactly when the numbers are going to hit.

I will say, as we get closer to into Q2, I think we'll probably -- I think there's a good chance we'll narrow our guidance and revise it. It's a little bit broad, having [ 38 to 42 ] and 6% to 9% on the cash flow side. But our plan would be to narrow that to the extent that we can during the second quarter.

Yes. And I know that it is March 11, and people are probably wondering like, what are you going to do first quarter. And what I will say is the good news about much of what we're doing today is it's not seasonal. Our intent is to do the best deals that we can and obviously, for the company, but for our partners because we believe that those will lead to additional.

We said double-digit increases in cash flow and top line revenue for the year that may seem a little conservative to people or a little lukewarm in light of the fact that we did 40% and 46%. But our intention as we give guidance is to beat that guidance. And that's the approach that we're taking, and we believe that over the year, that will that will yield the best results for us. Thank you for that question, Chris.

Clint, Brady, thank you. This is the end of the CuriosityStream Q4 and year-end 2025 earnings call. Thank you again to all of the participants on the line for staying with us through the technical difficulties. Have a nice evening.

Greetings, and welcome to the CuriosityStream Third Quarter 2025 Financial Results Conference Call [Operator Instructions] Please note, this conference is being recorded. I will now turn the conference over to your host, Tia Cudahy, CuriosityStream's Chief Operating Officer. Please go ahead.

Thank you, and welcome to CuriosityStream's discussion of its third quarter 2025 financial results. Leading the discussion today are Clint Stinchcomb, CuriosityStream's Chief Executive Officer; and Brady Hayden, CuriosityStream's Chief Financial Officer. Following management's prepared remarks, we will be happy to take your questions. But first, I'll review the safe harbor statement.

During this call, we may make statements related to our business that are forward-looking statements under the federal securities laws. These statements are not guarantees of future performance, but rather are subject to a variety of risks, uncertainties and assumptions. Our actual results could differ materially from expectations reflected in any forward-looking statements.

Please be aware that any forward-looking statements reflect management's current views only, and the company undertakes no obligation to revise or update these statements nor to make additional forward-looking statements in the future. For a discussion of the material risks and other important factors that could affect our actual results, please refer to our SEC filings available on the SEC website and on our Investor Relations website as well as the risks and other important factors discussed in today's press release.

Additional information will also be set forth in our quarterly report on Form 10-Q for the quarter ended September 30, 2025, when filed. In addition, reference will be made to non-GAAP financial measures. A reconciliation of these non-GAAP measures to comparable GAAP measures can be found on our website at investors.curiositystream.com. Unless otherwise stated, all comparisons will be against our results for the comparable 2024 period. Now I'll turn the call over to Clint.

Thank you, Tia. We delivered strong Q3 results. Revenue grew 46% year-over-year to $18.4 million, exceeding guidance. Adjusted free cash flow rose 88% to $4.8 million, and adjusted EBITDA improved by $3.4 million year-over-year. Our 3 complementary growth pillars, subscriptions, licensing and advertising are driving momentum and strengthening CuriosityStream's position at the intersection of knowledge, media and AI.

I'll briefly recap the underpinnings of Q3 and then look ahead to 2026 and beyond. Licensing revenue increased over 40% year-over-year, reflecting the strength of our team, demand for our corpus and the trusted relationships we built with traditional media partners and hyperscalers. We engaged with 9 key partners across video, audio, script and code and delivered over 1.5 million distinct assets.

To achieve dominance as a provider of AI training data, we've assembled nearly 2-million hour library of video and audio across multiple genres, content largely cannot be scraped from the open web. We've also expanded our large-scale data structuring and metadata capabilities so we can meet partners' volume requirements and bespoke specifications for high integrity and rich data sets.

In parallel, we broadened traditional content partnerships with leading global broadcasters, streamers and pay TV networks, including AMC, Netflix, Foxtel and a range of licensees across Asia. Overall subscription revenue, retail and wholesale combined was down year-over-year but increased sequentially every quarter in 2025. Importantly, our sequential growth in subscription revenue has been driven by daily operating focus, not simply by implementing price increases like many subscription services.

In Q3, all 3 of our subscription services launched with partners in key English-speaking markets, the U.S., Australia and New Zealand as well as in our non-English market, Germany. Extensions with partners like Amazon and new multifaceted agreements with partners like TMTG further support this growth trajectory.

While not yet a separate revenue pillar at scale, we continue to build on the solid foundation of our advertising business. Our U.S. Hispanic and flagship FAST channels recently launched on Amazon, Roku, LG and Truth+. We also launched a 2-hour branded block on Australia TV's free-to-air broadcast channel, an initiative we plan to replicate with additional partners.

Given the quality and volume of content we control, we see meaningful advertising and sponsorship opportunities across FAST, AVOD, social, pay-TV and free-to-air. To thoughtfully capture this opportunity, we plan to install a proven leader to run the business in early 2026. We are particularly proud that adjusted free cash flow increased 88% to $4.8 million this quarter. This reflects a focus growth strategy and a sustained commitment to rationalizing our cost base, especially hard, largely nondiscretionary costs. Cost discipline is a strategic advantage, one that supports pricing power, resilience and durable growth. Despite higher storage and delivery expenses from managing a large content library, we more than offset those increases through disciplined expense management.

Looking ahead, while we are not yet providing guidance for 2026, we expect overall subscription revenue, retail and wholesale to grow faster in 2026 than in 2025, supported by a strong launch pipeline and new pricing and packaging across our own services, including our premium tier. We anticipate high-growth licensing to continue and believe licensing will exceed subscription revenue in 2027, possibly earlier.

We expect significant year-over-year growth with existing partners and believe our roster of AI licensing partners could double or even triple in 2026. Beyond training, we see additional monetizable grants of rights becoming part of our agreements. Given the quality and scale of our corpus, which we expect to more than double in 2026, and our ability to structure and deliver data at scale, we believe we will solidify our position as the leader or among the top 2 or 3 video licensers for AI development.

In summary, we believe that we will continue double-digit growth in both revenue and cash flow, driven by our 3 complementary pillars: subscriptions, licensing and advertising, while continuing to improve efficiency. We intend to pay 2026 dividends from cash generated by operations as we did in 2024.

Our balance sheet remains strong with over $29 million in liquidity and no debt, giving us substantial flexibility. At today's share price, we're a growth company that also offers a dividend yield of well over 8%. It's an exciting time to be in the media business. Opportunities abound, and we intend to swarm them with discipline. Over to you, Brady.

Thanks, Clint, and good afternoon, everyone. Our full financial results will be in the 10-Q that we'll file in the next day or 2, but let me hit some of our third quarter highlights. As Clint said, in the third quarter, we reported revenue of $18.4 million, exceeding our guidance and a 46% increase compared to $12.6 million a year ago. Likewise, we reported another quarter of positive adjusted EBITDA, which came in at $3 million.

This was an improvement of $3.4 million from a year ago and essentially flat from last quarter, which was a record quarter for us. This was also our third sequential quarter of positive adjusted EBITDA. Adjusted free cash flow came in at $4.8 million, an increase of $2.3 million compared to last year. This also represented our seventh quarter in a row of positive adjusted free cash flow. Third quarter revenue was led by our subscription business, which came in at $9.3 million, a sequential increase.

Content licensing came in at $8.7 million in the quarter, an increase of over $7 million or 425% from last year, driven by continued growth in AI training fulfillments. Looking at our year-to-date numbers, licensing generated $23.4 million through September, which in perspective is already over half of what our subscription business generated for all of 2024. Third quarter gross margin was 59%, improving from 54% last year.

We continue to see reductions in noncash content amortization, although our distribution and storage costs have increased slightly due to licensing and acquisition of content through revenue share arrangements. Combined costs for advertising and marketing plus G&A were higher by 52% compared to last year. This increase was the result of a noncash charge for stock-based compensation of $7 million or about $0.12 on a per share basis.

G&A also included a number of onetime expenses associated with our August secondary stock offering. Were it not for the noncash SBC and the common stock sale, G&A would have declined in the quarter. We reported a third quarter net loss of $3.7 million or $0.06 a share. This compares to a $3.1 million net loss in the third quarter of 2024. While our revenue was up materially from last year, the net loss was driven by the onetime charges and noncash SBC. Were it not for these specific nonrecurring or noncash charges, we would have posted our third quarterly net income this year.

And as we said earlier, adjusted EBITDA was $3 million in the third quarter compared to a loss of $0.4 million a year ago. Adjusted free cash flow was $4.8 million in the quarter compared with $2.6 million a year ago. And through the first 9 months of 2025, adjusted free cash flow was $9.6 million, which is more than the company generated for all of last year. In September, we paid our regular $4.6 million dividend, and we ended the quarter with total cash and securities of $29.3 million and no outstanding debt.

We believe our balance sheet remains in great shape. Based on yesterday's share price, CuriosityStream is generating an adjusted free cash flow yield of over 7% and a current dividend yield of over 8%. Also just after quarter end, on October 14, 6.7 million of our warrants expired unexercised. While these warrants have been trading well out of the money for some time, this expiration of all of the company's outstanding warrants reduces potential dilution and should eliminate any lingering share overhang associated with these instruments.

Looking ahead, for the fourth quarter, we expect revenue in the range of $18 million to $20 million, which would imply full year 2025 revenue in the range of $70 million to $72 million or a 38% to 42% revenue increase from 2024. We expect fourth quarter adjusted free cash flow of $2.5 million to $3.5 million, which would imply full year 2025 adjusted free cash flow of $12 million to $13 million or a 27% to 37% free cash flow increase from 2024.

We're not yet providing guidance for 2026, but we believe that our top line and bottom line growth will continue into next year. And although we're obviously using some of our cash and investment reserves to pay our dividends in 2025, we intend to fully cover our 2026 dividends from operating cash as we did in 2024. With that, we can hand it back to the operator and open the call to questions.

[Operator Instructions] And our first question will come from Laura Martin with Needham & Company.

So Clint, a strategy one for you first, and that is -- so I know you've been a media CEO for a long time, but the returns on capital in this new revenue stream of licensees are like 10x higher. So what I don't understand is why are we taking these fabulous revenue and investing and hiring a guy to do media in Australia, which is offshore, lower margin, lower returns on capital.

Why don't we just stick with -- stick with focusing on becoming sort of the go-to de facto AI guys that are independent and put all -- say no to media stuff, which is our path. Why are we adding stuff that's lower return on capital just because that's where we came from. Let's start with that one.

I appreciate the question, Laura. You cut out a little bit. The Australia reference was -- I just -- I didn't hear what you're referring to in Australia. I think I got the base...

You're hiring, right? What's the new guy going to do when you hire in?

We haven't hired anybody new. We just -- we announced a promotion of one of our guys who has been focused on helping to craft our AI relationships. We announced that last week, if that's what you're referring to.

No, I thought you said on the call that you were going to...

Okay. Yes, I did say on the call that we're going to hire a sales leader, yes. And let me take the -- and I think your question is a good one as it -- and so I think we've done a great job certainly on the cost side here, and we've done a really good job across the company, getting this business rolling. But we do need some additional sales leaders, even some who are really seasoned. And so there's an opportunity for us to do that.

And sometimes, if you get a chance to bring somebody into the mix, it's a little bit like the NFL draft. We're not necessarily drafting for pure position, but if you can bring on an A+ player with real talent, you take the opportunity to do that. And so I didn't mean to imply that he would be -- or she would be working in only one area, but really helping on the revenue generation side.

That is a -- key for us is we do need some help there. And we've done, as I said, I feel like we've done a nice job on the cost side, and we've laid some good groundwork so that if we bring in a couple of really strong players, it can have an accelerating impact on what we're doing. Let me do dodge, hopefully, I addressed that. If I didn't, Laura, ask me to clarify.

Well, I just want to be sure we're staying in the AI business is not...

Yes. 100% Yes, yes. Absolutely, Laura. Absolutely. And yes, and I know that you've expressed some concern in the past about smoothing out the revenue. And I'd love to address that if that's still a question on...

That was my second question, which is...

Okay. You're not the only one asking. So look, there will always be some lumpiness in licensing, but we're going to smooth it out over time. And we're going to accomplish this, and we are accomplishing this by both operational and contractual means. So operationally, as we increase our roster of partners, we reduce lumpiness. We believe that we'll double or triple our number of partners in the AI licensing area by the end of 2026.

And in 2027, possibly earlier, as more open source models become accessible, there will potentially be hundreds and even thousands of companies who will need video to fine-tune specific models for consumer and enterprise purposes. Some refer to this as the open source and tune evolution. Contractually, structurally, SaaS, something you're familiar with, of course, Software as a Service. We know that's beloved by the software industry. We know that is beloved by investors.

So with the type of volume we control, we've had discussions around structuring certain agreements as CaaS or C-a-a-S or Content as a Service, where we grant access over a term, so as a subscription with access to clouds of content with lots and lots of hours. Now in these deals, we need to make sure we have proper minimums in place and a few other safeguards, but this is a proven model that I think a lot of people love and that will enable smoother quarters and tighter predictability over time.

And our next question comes from Jason Kreyer with Craig-Hallum.

Great job, guys. So maybe kind of building on just the library and the AI opportunity. Clint, just curious if you could talk about how AI licensing has evolved over the last year. You had mentioned 9 partners this quarter. Just maybe talk about how broad is the demand for your corpus and how frequently are those platforms coming back to get more and more of your content?

Yes. So we've done about 18 fulfillments to date, and we've done that across 9 partners. And so without naming names, based on the math, I think we can represent that we've done 2 or 3 more renewals with some key existing partners. And as we look out into 2026, we see -- we suspect really the revenue from our existing partners, they'll probably comprise 60% to 80% of the AI licensing revenue and 20% to 40% will come from new partners.

So we do see the roster increasing significantly. And we also see real opportunity for licensing beyond simply a training right, additional grants of rights like display rights or transformative rights or adaptation rights or even certain derivative rights or possibly even some that are as of yet unnamed. I mean we're building long-term relationships, and we're committed to making sure that as we enter into all of these agreements, it's not one and done.

And so I think from an -- and then to answer your question about the evolution and the changes, I think if you look at the first deals that we did, it was just get people finish content, let them use that for training the models. Today, obviously, people are still looking for finished content, looking for raw video. But there's a real advantage to being able to structure the data in a way that many of your competitors can't.

And what I mean by that is just the ability to clip content to index content to annotate and to then deliver it in 7- to 20-second clips with really detailed enriched metadata. So one thing I didn't mention on the call is in Q3, we entered into some of the highest cost per hour or cost per minute agreements by far that we had heretofore not entered into. So that's a function of being able to create things that are a little bit more bespoke and a much enhanced ability on our end to structure a data.

And Clint, when you talk about having nearly 2 million hours in the library, how does that split between what's available for AI licensing, what's available for streaming by consumers? Or is that the same number for both?

Yes. No, it's not the same number. It's a good question. The overwhelming majority of that is for AI licensing. So we have a lot of content partners, probably over 150 different content partners for our subscription services and for our ad-supported services.

And obviously, as we're building our AI library, so those are some of the first people that we went to. But the overwhelming majority of that is for AI licensing. We are increasing our volume of rights in our traditional platforms, but the overwhelming majority is for AI licensing.

You've nearly doubled that in like the last quarter or 2 or doubled the library. I'm going to assume that's a proxy for effectively doubling the AI library over the last quarter or 2. Curious, is that an indication of, hey, you guys are getting better at figuring out what these AI platforms need and you're going out and sourcing a lot more of that? Or is there a different reason that we're not thinking about it and why you're adding so much more content to the library?

Well, you're 100% right on the first part. We are sourcing specific content. So that's part of it. At the same time, we feel like one of our advantages is the fact that we have existing relationships with so many production companies, so many distribution companies, so many creators, so many people who own and control large libraries of content that wherever it makes sense, we want to put our foot on the gas.

And we're not going to be the #1 subscription service in the world. I mean Netflix and Amazon, they've got escape velocity. They have that covered. However, we believe that we can be, if not #1, one of the top 2 or 3 licensors of video for AI training and other purposes. And part of the way that you do that is you try to generate some escape velocity on your own. So as we build out the volume of our library, it makes us more appealing even more so as a one-stop shop for any of our partners.

And moving on to David Marsh with Singular Research.

Congrats on the quarter. I wanted to start, Brady, if I could. Could you give us a little bit more explanation of the stock-based comp here in the quarter? It's -- I know you had some comments about it in your prepared remarks, but I'm still a little bit confused by it. So I was just hoping you could maybe just give us a little bit more color.

Yes, sure. So -- and the 10-Q will be out either tomorrow or the next day, possibly Friday, but a lot of the details from that will be -- for the stock-based comp will be in that document. But a number of employees received a market-based SBC warrants and awards during the quarter, during -- actually in July. You'll see some of those in the Form 4s that the executives filed and a number of other employees received those as well. The market-based components of those were something new this quarter.

And because of that, we had to take a much higher grant date fair value than we would have if the awards had just been purely time-based or purely internal performance-based. Because the stock -- because of the stock market component of those awards, the market-based component, we have to value those differently. So I think the total value -- and this will be in the Q, but it's well -- it's into the 8 figures, the total value of all of those awards, and that has to be expensed over an aggressive period of time, more aggressive than if they were purely time-based over 4 years.

So that's why we are -- that's why we had the unusually high SBC in this quarter. And you'll see we have another -- we disclosed an amount that will need to be expensed over the next several quarters. So hopefully, that will be easier for you guys to factor into your models. Is that helpful, Dave?

Yes. That's really helpful. I appreciate it. Yes, it looks like almost $7 million a quarter. A couple of follow-up questions on that. How is that reflected in the current diluted share count, if it is at all? And how will it impact the share count going forward? And then just again, around SG&A, I mean, would you expect it to retreat pretty substantially in Q4? Is this an annual like accrual type thing? Or is this recurring quarterly?

Yes. So if you do the math and you look at all of our public filings, we -- I'd say it was a majority of what we will do in a year were granted in the quarter. I wouldn't expect there to be anything to that level in the next several quarters. I don't know exactly what our SBC will be for Q4, but I think the majority of what we will do for this year's grants we had to expense in Q3.

Okay. That's really helpful. Turning kind of more strategically, Clint, thanks for the color around the licenses. It's helpful. Could you talk about content or subscriptions a bit though? You guys have launched a number of new -- kind of new markets over the last several quarters. And can you just give us an update on reception and what kind of momentum you're seeing in some of those new markets and new platforms, please?

I appreciate that question. And if I may, if I could just add a little color to what Brady shared. I want everybody to know that all of our employee equity grants at this point, including or especially mine, are linked to financial performance of the company, all quantitative goals. So as a company, we're very tightly aligned around business and performance compensation. It's a key pillar of culture, and it's a key pillar of our compensation structure.

We are not a bloated media company where people claim to fat-based salaries and guaranteed bonuses really strive to be the farthest thing from that as a performance-based company. So equity is a critical component of this pursuit, and it's also a way to help maintain a competitively advantageous operating budget. But getting back to your question around subscriptions.

We have had a number of new launches with partners like Amazon in Australia, in New Zealand. And as they roll out subscription-based services around the world, we want to roll out with them. It's a -- building a subscription business is something where there are huge barriers to entry. And so for us, like beyond just the reliable recurring revenue of our subscription services, it's a moat and it's a competitive advantage in our licensing acquisition efforts.

And as a result of these deals, David, and as a result of our sort of pipeline visibility, we're supremely confident that our overall subscription revenue, retail and wholesale will grow at a higher rate in '26 than in '25. And again, we're confident because we have visibility in the third-party pipeline, meaning new and meaningful wholesale distribution agreements, which will kick in as well as channel store launches inside and outside the U.S. for our SVOD services and also because we're diligently planning and taking the requisite steps to execute new pricing and packaging in 2026.

I mean if you look at a lot of public and private companies in the subscription space, the way that they've grown is through simply or I might add often just lazily raising price. That's something we have the capacity to do, but we're trying to get to the core of what we really need to do based on our marketing spend to grow subscribers.

And -- let me jump back in. I think I forgot to answer your question on dilution. A portion of the RSUs for Q3 already vested in those were those did factor into our share count for dilution purposes. Obviously, we reported a net loss for the quarter. So it was overall anti-dilutive. But going forward, assuming we're -- for net positive quarters, we'll certainly have to include those -- all of those are fully diluted.

And Patrick Sholl with Barrington Research has our next question.

Just first one on the guidance in the quarter. Could you maybe just talk about the free cash flow guidance? It seems like relative to the increase in revenue, there's kind of limited free cash flow growth in the quarter. Is that mostly just a timing issue? Or is there anything to keep in mind there?

Timing issue.

Okay. And then on the content library for AI licensing. Can you just maybe talk about like the different markets between the content that you have that is from maybe your partners that also work with you on the streaming side versus other partners? And just maybe the different dynamics on margins and use cases and partners for licensing there.

Yes. It's a good question, Pat, and thank you. So when we were first building our library for AI licensing, we went to people with whom we had great existing relationships. And most of those companies are in the factual space, nonfiction entertainment, science, technology, history, travel, lifestyle, et cetera.

In an effort to try to generate, again, escape velocity or dominate in the space, we then, at the same time, or shortly thereafter, start reaching out to people that we knew distributors who controlled content in other genres, general entertainment, sports as an example. And so with that type of corpus and with the way that we're able to structure some of that data, that's made a real difference as it relates to our overall proposition.

I mean people love our content and love working with us because we have diversity of content, we have high quality, and we're able to just enrich the data in a way that's unique. But we do have some content whereas a lot of the content that we have in our AI corpus, we could put on our services, there is some content that goes well beyond the factual content that you would see on any of our subscription services or our ad-supported services.

And ladies and gentlemen, thank you for your participation. This does conclude today's teleconference. You may disconnect your lines, and have a wonderful day.