Funding Round
<cite index="5-1">Baseten, an artificial intelligence (AI) inference company, announced a $1.5 billion Series F financing led by Altimeter Capital, Conviction, and Spark Capital.</cite> <cite index="1-4">Sands Capital and Wellington Management participated as co-leads.</cite> <cite index="1-5">Other investors include IVP, 01A, Greylock, Durable Capital Partners, Blackbird, Verified Capital, Battery Ventures, and D. E. Shaw Ventures, alongside existing shareholders.</cite>
<cite index="2-4">The round includes investments made across two tranches at valuations of $13 billion and $11 billion respectively, and reflects surging demand for inference at the application layer as closed-source and open-source models converge in capability, cost, and customization.</cite> <cite index="4-9">The company has now raised more than $2 billion to date.</cite> <cite index="7-1">The raise is corroborated by filings with the U.S. Securities and Exchange Commission (SEC), which recorded $1,500,000,000 in new Series F funding.</cite>
Company Background
<cite index="5-5">Founded in 2019 and based in San Francisco, Baseten counts among its backers Altimeter Capital, Battery Ventures, Blackbird, BOND, CapitalG, Durable Capital Partners, Greylock, IVP, NVIDIA, Sands Capital, Spark Capital, Verified Capital, and Wellington Management.</cite> <cite index="5-3">The company builds systems software that runs the entire workload for AI applications—from Graphics Processing Units (GPUs) and autoscaling to observability, billing, and developer tools—so teams can focus on models and user experience rather than infrastructure.</cite>
<cite index="12-2">Baseten monetizes usage-based AI inference, charging customers either for API consumption on open-source models or for minutes and hours of GPU capacity.</cite> <cite index="5-4">Its customers include leading AI companies such as Abridge, Clay, Cursor, Lovable, Mercor, and OpenEvidence, among others building specialized models for their domains.</cite>
Market Context and Growth Metrics
<cite index="11-3">Post-trained open-source models are increasingly delivering frontier-level performance at a fraction of the cost, and leading application-layer companies now direct 30 to 50 percent of model spend toward custom and post-trained models.</cite> This dynamic has elevated demand for independent inference infrastructure, with enterprises moving away from sole reliance on closed-source Large Language Model (LLM) Application Programming Interfaces (APIs).
<cite index="11-6">Revenue has grown approximately 20 times year-over-year, and the platform now processes more than 1 billion inference calls every day—across 87 clusters globally and spanning 18 clouds, a multi-cloud architecture that is itself a core part of what customers are buying.</cite> <cite index="18-4">In the first quarter of 2026, Baseten tripled its annualized revenue from $200 million to $600 million in a single quarter.</cite>
Prior Fundraising History
<cite index="1-6">Prior to this round, in January 2026, Baseten secured $300 million in a Series E led by anchor investors IVP, CapitalG, and NVIDIA, valuing the company at $5 billion.</cite> <cite index="9-1">The Series F marks the company's fourth capital raise in 18 months.</cite> <cite index="6-3">Baseten's Series D, raised in September 2025, valued the company at $2.15 billion, itself more than doubling the $825 million valuation set during the $75 million Series C in February 2025.</cite>
Planned Use of Capital
<cite index="3-6,3-7">The new capital will be used to expand the company's workforce, increase computing capacity, and accelerate enterprise go-to-market initiatives. Baseten plans to triple its headcount this year as it scales engineering, research, operations, and commercial teams to support growing customer demand.</cite> <cite index="9-7">The company plans to invest the capital in compute infrastructure, software, and engineering talent as more enterprise customers move from closed APIs toward open-weight models that can be customized and optimized for specific workloads.</cite>