Baseten is close to finalizing a $1.5B funding round at a $13B valuation, according to a Wall Street Journal report, capping a stretch of fundraising velocity that has few peers in the AI infrastructure market. The round arrives just five months after the company announced a $300M Series E at a $5B valuation, and nine months after that round, the company had raised a $150M Series D. If the new financing closes on those terms, Baseten's valuation will have climbed 160% in less than half a year.
Spark Capital, Sands Capital, Altimeter Capital, and Wellington Management are co-leading the deal. The investor syndicate represents a mix of crossover and growth firms — the kind of capital base that typically lines up around a company preparing for the public markets, not one still iterating on product-market fit.
The round is split-priced, a structure the Journal flagged explicitly. Some investors are coming in at the $13B headline valuation, while others are entering at $11B. Split pricing has become a recurring feature of late-stage AI deals: it lets lead investors anchor a high headline number that benefits founders and existing shareholders on paper, while giving other participants a more conservative entry point. The optics are cleaner than the underlying economics suggest.
Key facts
- 01Baseten is close to finalizing a $1.5B round at a $13B valuation, per WSJ reporting.
- 02The deal comes just five months after a $300M Series E at a $5B valuation.
- 03The round is split-priced: some investors enter at $13B, others at $11B.
- 04Spark Capital, Sands Capital, Altimeter Capital, and Wellington Management are co-leading.
- 05The Series E followed a $150M Series D raised nine months earlier.
Baseten, founded in 2019, sells inference infrastructure — the layer that runs an AI model after a user submits a prompt. Its pitch is speed and cost control: route each request to the model best suited for the task, lean on competent open-source alternatives when they suffice, and avoid burning frontier-model dollars on workloads that don't need them. That value proposition has resonated as enterprise AI bills compound.
The category is on fire. The Next Wave has called the current moment an "inference gold rush," with venture capital flooding into companies building the layer between models and end users. Training got the early hype and the early dollars; inference is where the recurring cost — and the recurring revenue — actually lives. Every prompt is a transaction, and the providers that can shave latency and cost off each transaction are positioned to capture a slice of every AI product in production.
Baseten's trajectory through 2025 and into 2026 reflects how aggressively investors are pricing that thesis. The $150M Series D, the $300M Series E at $5B, and now a $1.5B round at $13B amount to roughly $1.95B raised across three rounds in under two years. Few infrastructure companies have ever compounded paper value at that rate without a public listing in between.
Comparable activity in the broader market underscores the pattern. AI General Intuition is in talks to raise $300M at around a $2B valuation, a story we covered separately this week, and that deal is one of several mid-stage AI rounds being priced well above traditional revenue multiples. Capital is concentrating in the companies that sit closest to model deployment.
There are reasons for caution embedded in the structure itself. Split-priced rounds can mask the true clearing price of a company's equity, and a 160% step-up over five months sets a bar that subsequent rounds — or any eventual IPO — will have to clear. If the inference market commoditizes faster than expected, or if hyperscalers like AWS, which is now in talks to sell its Trainium chips to outside buyers, undercut independent inference providers on price, the gap between the $11B and $13B tranches could look like the more honest signal in retrospect.
For now, Baseten is one of the clearest beneficiaries of where AI spending is actually flowing. Training runs grab headlines, but the bills that enterprise CFOs are starting to scrutinize are inference bills, and the vendors that can route workloads intelligently across models — including cheaper open-source ones — have a structural advantage. A $13B valuation is a bet that Baseten can hold that position as the market scales by an order of magnitude, against well-capitalized competitors and the cloud incumbents who would prefer to own the layer themselves. That bet is now $1.5B larger.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
