Gimlet Labs now has a valuation of $3 billion, after the two-year-old AI startup raised $300 million in a Series B led by Andreessen Horowitz, the company said Friday.
Key takeaway
Gimlet Labs has raised $300 million in Series B funding, led by Andreessen Horowitz, bringing its valuation to $3 billion
The platform distributes AI inference workloads across chips from six manufacturers: Nvidia, AMD, Intel, Arm, Cerebras and d-Matrix
A former Bitcoin mining player turned to cloud infrastructure for AI closed a $3 billion funding round last month, at a $30 billion valuation
Gimlet does not disclose either its customer count or its revenue
This round finances what Gimlet presents as the sector’s first "multi-silicon inference cloud": a platform capable of routing AI inference workloads across chips from six providers — Nvidia, AMD, Intel, Arm, Cerebras, and d-Matrix, according to the press release. The software lets developers write their inference code once, then run it on the chip that is cheapest or fastest at any given moment, without locking customers into a single vendor.
How Gimlet Labs' valuation climbed to $3 billion
This valuation propels Gimlet into the small circle of AI infrastructure players posting multibillion-dollar figures, without designing their own chips.
Gimlet’s central argument rests on "agentic AI": systems that chain together multiple inference steps, call external tools, and make certain decisions without human validation. These workflows shift unpredictably between short bursts of computation and long chains of reasoning.
An architecture centered on a single type of chip ends up wasting resources on the stages for which it is not optimized. Standing above all chips, rather than betting on a single architecture, is the product bet that the $3 billion valuation must now justify.
The rush to fund AI infrastructure is nothing new
Gimlet’s raise is part of a series of massive funding rounds in AI infrastructure this year.
Also read: Polygon opens a crypto donation route to the relief fund for the Nepal disaster
A Bitcoin (BTC) mining company turned to cloud infrastructure for AI finalized a $3 billion funding round last month, bringing its valuation to $30 billion. A clear sign of how quickly capital is shifting away from mining farms and toward GPU clusters. Gimlet’s valuation is about one-tenth of that amount, a gap that also reflects the difference in scale between the two rounds.
The multi-chip bet still has to prove itself
Gimlet does not disclose either the number of its customers or its revenue level.
The $3 billion valuation therefore rests above all on an expectation of future demand, far more than on current performance, and it comes with a structural risk: the company depends on Nvidia, AMD, Intel, and Cerebras being willing to let a third party orchestrate their chips, rather than integrating that software layer themselves. If any one of the six partners tightens access, the promise of "run anywhere" would immediately lose some of its appeal.
Coming up: OpenAI launches GPT-6 Astra, its most advanced model to date