Baseten raises $1.5B at $13B valuation for AI inference

Baseten is on the verge of closing a $1.5 billion funding round that pushes its valuation to as high as $13 billion, marking one of the largest private raises in the AI infrastructure space.

The mechanics of the deal are drawing as much attention as the dollar figure. The round operates on a dual-tiered structure, with some investors entering at an $11 billion valuation and others paying up for a $13 billion price tag, according to the company’s disclosure to the Wall Street Journal. This layered approach has become a common tactic among AI startups aiming to inflate headline numbers while accommodating different risk appetites. The financing is co-led by Altimeter Capital, Conviction, Spark Capital, Sands Capital, and Wellington Management.

Baseten operates in the less glamorous but increasingly vital layer of the AI stack. Founded in 2019 and based in San Francisco, the company provides the software and computing capacity businesses need to run inference,the stage where a trained model actually processes a query. Rather than building its own data centers, Baseten rents capacity from roughly 20 cloud providers and layers its proprietary inference software on top, allowing customers to deploy and fine-tune models using their own data.

Much of that workload runs on cheaper, open-source models rather than the frontier systems from OpenAI or Anthropic. That is the core bet. “Open-source models are getting very, very good,” CEO Tuhin Srivastava told the WSJ, noting that customers increasingly mix open and proprietary models depending on the complexity of the task.

Investors are re-rating that thesis at breakneck speed. Baseten was valued at $5 billion as recently as January, when it raised a $300 million round backed by IVP, CapitalG, and Nvidia. The new deal more than doubles that figure in roughly five months. As recently as September 2025, the company was worth about $2.15 billion.

The driver is surging demand. Customers including Cursor, Mercor, and OpenEvidence use Baseten to serve models, and the company reports that some have cut costs sharply by shifting workloads to open-source options. One client reportedly runs a task at about 30 percent of the cost of a proprietary alternative.

The raise lands in the middle of a full-blown AI inference gold rush, where the layer that makes models run efficiently has become some of the most sought-after infrastructure in the industry. Cerebras, an inference-chip maker, went public earlier this year, and Fireworks AI is chasing the same model-serving market.

It also lands in the middle of a price war. Many of the most-used open-source models now come from China, including DeepSeek and Moonshot AI, and even Nvidia has released an open family called Nemotron. With OpenAI reportedly weighing steep price cuts to keep up, that pressure is bad news for the frontier labs and, in theory, good news for whoever helps companies run the cheaper options.

The risk sits in the same place as the opportunity. Baseten is being valued as a winner of a cost collapse, even as that collapse squeezes margins across the field. A $13 billion price tag on a company at the center of a price war is, itself, a bet that the shovels outlast the gold.