AI inference startup Baseten is reportedly close to finalizing a new mega-round of funding, according to TechCrunch.
The report says Baseten is raising $1.5 billion at a $13 billion valuation. Notably, this would come just months after the company's last large fundraise — a sign of how quickly money is moving into this corner of the AI industry.
TechCrunch frames the deal as part of what it calls an ongoing "inference gold rush." Inference refers to the stage where a trained AI model actually does its job — generating answers, images, or predictions in response to real user requests. Every time someone queries a chatbot or runs an AI feature, that's inference at work.
As reported, the figures here — the $1.5 billion round and the $13 billion valuation — are not yet final. TechCrunch describes the deal as close to being finalized rather than confirmed, so terms could still shift.
Why it matters: much of the public attention on AI has focused on training ever-larger models, but running those models at scale for millions of users is its own enormous, costly business. Investors pouring billions into a company like Baseten signals a bet that the real, recurring demand — and money — in AI lies in serving models efficiently, not just building them.