Latest / The Startup Exit Podcast with Fexingo: IPOs, Acquisitions, and Founder Liquidity Events / How OpenRouter Doubled Valuation in a Year Without VC
Transcript
- Lucas: So OpenRouter — the AI inference marketplace — just raised at a one point three billion dollar valuation, more than doubling in a year. And the interesting part isn't even the number. Luna: What is it then? Because that's a pretty big number for a company that most people haven't heard of outside developer circles. Lucas: Right. The interesting part is how they got there. They didn't take traditional VC money early on. They bootstrapped, then used a strategic round from some pretty specific investors. It's a playbook that a lot of founders are starting to pay attention to. Luna: Especially in this market. We just saw RIVN up twelve percent in five days, ARKG up eleven percent — there's clearly appetite for growth stories, but the IPO window is still narrow. Lucas: Exactly. So OpenRouter's model is essentially a marketplace for access to different large language models. Developers pay per token, and OpenRouter routes queries to the cheapest or fastest provider. It's infrastructure, not a consumer product. Luna: And the valuation jump from roughly six hundred million to one point three billion in one year — that's a roughly one hundred percent increase. What drove it? Lucas: Revenue growth, mostly. AI inference demand is exploding. Every startup building on top of GPT or Claude or Llama needs a way to manage costs and latency. OpenRouter sits in the middle and takes a cut. Their revenue multiple probably contracted slightly, but the absolute revenue number went way up. Luna: So they're not just a hype story. There's actual usage behind it. Lucas: Exactly. And that's what made them attractive to strategic investors — the kind who aren't just writing a check and hoping for an IPO. They brought in investors who could also be customers or partners. That changes the exit calculus. Luna: How so? If you take strategic money, doesn't that limit your exit options? You can't sell to a competitor of your investor. Lucas: It does, but it also creates a more predictable path. A strategic investor might be the acquirer themselves, or they might help you find a buyer that fits. And in a market where IPOs are still uncertain — look, we've seen how Stripe used private secondaries to give employees liquidity without going public — strategic rounds can be a way to extend the private runway indefinitely. Luna: So OpenRouter might never IPO? Or at least, they don't need to. Lucas: They might not. The founder has said they want to stay independent, but at one point three billion, they're big enough that an acquisition would be meaningful. The question is who buys them. Cloud providers like Amazon or Google could want the routing layer. Or a telco. Or a chip company wanting to own the software stack. Luna: And if they stay private, they can keep using secondaries to let early employees cash out. We covered that a few episodes ago — how secondaries have become a liquidity lifeline. Lucas: Right. And OpenRouter is a perfect case study. A company that's growing fast, in a hot space, but doesn't want to deal with the quarterly pressure of being public. They can use strategic rounds and secondary sales to give people exits without an IPO. Luna: Let's talk about the specific round. Who led it? Was it a traditional VC or a corporate? Lucas: It was a mix. A few existing investors participated, but new investors included a large cloud provider — unnamed, but you can guess — and a couple of sovereign wealth funds. No traditional VC led the round. It was all strategic capital. Luna: That's a signal. Sovereign wealth funds are long-term. They don't need a five-year exit. They're okay holding for a decade. Lucas: Exactly. So OpenRouter's cap table now has investors who are aligned with a longer timeline. That gives the founder leverage. They don't have to sell or IPO until the timing is right for them, not for a fund's life cycle. Luna: And the valuation itself — one point three billion — is it fair? Or is it frothy? Lucas: It depends on what multiple you assign. If they're doing, say, fifty million in annualized revenue, that's a twenty-six times multiple. For a company growing one hundred percent year-over-year, that's not crazy. But if growth slows, the multiple compresses fast. We saw that with a lot of the 2021 vintage unicorns. Luna: So the risk is that they're priced for perfection. One bad quarter and the next round could be a down round. Lucas: Right. But the structure of the round — strategic investors, no traditional VCs — means the downside might be softer. Strategic investors care about the technology, not just the return. They might inject more capital at a lower valuation just to keep the company independent. Luna: Is there a precedent for this? A company that used strategic rounds to stay private and then eventually exited on their own terms? Lucas: Sure. Palantir stayed private for years with a mix of government and strategic investors. They eventually went public, but on their own timeline. And more recently, companies like Databricks have used private secondaries to give employees liquidity without an IPO. The path is becoming more common. Luna: And for founders listening, what's the takeaway? Should they avoid VC? Lucas: Not necessarily. VC money is still the fastest way to scale if you need to burn cash to win a market. But if your business has good unit economics and you can grow without massive capital, strategic rounds give you more control. You just have to be willing to trade some optionality — you can't sell to a competitor of your investor. Luna: And you have to build a business that strategic investors actually want. Not every startup is infrastructure for AI. Lucas: Right. OpenRouter is in a unique position because they sit between the model providers and the end users. That's a natural place for a cloud provider to want to own. But if you're a consumer app, the strategic interest is different. It's about user data or distribution. Luna: So the lesson is: know who would want to own you, and structure your funding accordingly. Lucas: Exactly. And OpenRouter seems to have done that really well. Their cap table now reads like a coalition of stakeholders who all benefit from the company's independence. That's a strong position heading into whatever comes next — whether that's an IPO, an acquisition, or staying private forever. Luna: It's also a reminder that the exit landscape is more varied than people think. IPO isn't the only game in town. Lucas: And on that note — we talk about these strategies because we think they're useful for people building companies. If these conversations are useful for what you're building or running, we should mention that we deliberately don't run ads on these shows. Luna: Right. It's a choice. We'd rather keep the content clean and focused on what matters. Lucas: So if you want to support that choice, the link is buy me a coffee dot com slash fexingo. No pressure, just a way to keep it ad-free if you find value here. Luna: And we appreciate everyone who does. It lets us keep digging into stories like OpenRouter's. Lucas: So back to OpenRouter — the other thing worth watching is whether this model starts to get copied. If you're a founder in AI infrastructure, do you try to replicate their path? Luna: I think you have to. The bar for going public is still high. The IPO market in 2026 is better than 2023, but it's not 2021. Strategic rounds and secondaries are becoming the default path for companies that don't need to raise massive amounts of cash. Lucas: And that's a structural shift. We're moving from an ipo or bust mentality to a more flexible exit environment. OpenRouter is just one data point, but it's an instructive one. Luna: Especially because they did it in a year. From six hundred million to one point three billion. That's a lot of value creation for a small team. Lucas: It is. And it shows that in AI, the infrastructure layer is where the reliable value is right now. The application layer is crowded and volatile. But the plumbing — that's sticky. Luna: So if I'm a founder building an AI app, do I worry that OpenRouter becomes a gatekeeper? Lucas: You might. If they become the default way to access models, they have pricing power. But that's a problem for later. For now, they're still focused on growth and reliability. The gatekeeper question only matters if they succeed. Luna: Fair. And at one point three billion, they're succeeding. Let's see where they are in another year. Lucas: Yeah. That's the episode. Thanks for listening.