Baseten Reportedly Nearing $1.5B Round at $13B Valuation

Five months is not a long time between funding rounds at this scale. As TechCrunch reports, citing the Wall Street Journal, AI inference startup Baseten is nearing completion of a $1.5 billion round that would value it at $13 billion, a 160% jump from the $5 billion valuation attached to its $300 million Series E just five months ago. The round is reportedly co-led by Spark Capital, Sands Capital, Altimeter Capital, and Wellington Management.

There is an unusual wrinkle in the structure: some investors are entering at the headline $13 billion valuation while others participate at $11 billion. That pricing split is worth pausing on. It could reflect different risk tolerances among the co-leads, or a negotiation that ended with some parties unwilling to pay the top price but unwilling to walk away either. Either way, it is a sign of the intensity of competition to fund AI infrastructure right now, not straightforwardly a signal of weakness in the company.

Founded in 2019, Baseten operates at what observers are already calling "the inference gold rush," the race to build efficient infrastructure for the computational work that happens after a user submits a prompt to a language model. Its approach focuses on routing inference requests toward optimal models, with particular emphasis on cost-effective open-source alternatives to pricier commercial options from companies like OpenAI and Anthropic.

The honest caveat is that "nearing completion" is not closed, the reporting rests on a single outlet, and Baseten has not confirmed the terms publicly. What the story does not give you is any picture of Baseten's actual revenue or growth trajectory, or how the $11 billion and $13 billion tranches break down by investor and size, which matters a great deal for understanding who is betting what and at what risk.

For enterprises trying to manage inference costs at scale, a well-capitalized Baseten routing workloads across open-source model options is a meaningful counterweight to locking into a single commercial API. The open-source model ecosystem benefits too: more Baseten volume means more real-world demand for capable, affordable models, and that dynamic only strengthens if the round closes on these terms.