Baseten’s $1.5B raise shows the inference gold rush is far from over

Baseten is apparently at it again. The AI inference startup is reportedly closing in on a $1.5 billion funding round at a $13 billion valuation, according to sources who spoke with TechCrunch. That’s a massive number by any standard, but what’s wilder is that this comes just months after their last mega-round.

If you’ve been watching the AI infrastructure space, this shouldn’t come as a complete shock. The “inference gold rush” is real, and Baseten has positioned itself as one of the key picks-and-shovels providers for companies that need to run AI models in production. They’re not training models; they’re making sure those models actually work at scale when real users hit them.

What I find interesting is the timing. Most startups would take a breather after closing a large round, but Baseten seems to be sprinting. That tells me either they’re burning cash faster than expected to capture market share, or the demand for inference infrastructure is outstripping even their own projections. Given how many companies are now shipping AI features, I’d bet on the latter.

The valuation jump is telling too. If this round closes at $13B, that’s a significant step up from wherever they were in their previous round. For context, that puts them in rarefied air among AI infrastructure companies, competing with the likes of CoreWeave and Lambda in the cloud GPU space, though Baseten’s focus on inference specifically gives them a different angle.

Of course, $1.5B is a lot of money to raise in any environment, let alone one where investors are getting more selective about AI bets. The fact that they can pull this off suggests their revenue growth and customer adoption metrics are strong enough to justify the premium. I’d love to see their actual numbers, but we’ll have to wait for the official announcement or their eventual S-1 filing.

One thing worth watching: how this affects the broader inference market. Baseten’s aggressive fundraising puts pressure on competitors like Replicate, Modal, and even the cloud hyperscalers who offer their own inference services. It also signals that the market believes inference will be a massive, sustained business, not just a temporary spike driven by ChatGPT hype.

If you’re building AI applications today, the infrastructure landscape is getting more interesting by the month. More capital means more features, better pricing, and hopefully more reliability. But it also means the window for new entrants is closing fast. Baseten is making a bet that they can be the default platform for inference, and with $1.5B in fresh capital, they’ve got the ammo to try.