TypeSafe AI has secured $870 million in funding at a $7.5 billion valuation, following the breakout success of its Jev model, which became widely popular within weeks of its debut. The financing round was led by Andreessen Horowitz, with Sequoia and existing backer DCVC also participating.

The scale of the raise is hardly shocking given how quickly Jev took off after its Sept. 15 launch. According to the startup, one-third of Fortune 500 companies are already using the model — an unusually fast rate of enterprise adoption.

Jev relies on a transformer architecture, yet it is not a large language model. Rather than generating text, it produces probabilities — what the company describes as "calibrated decisions."

What has drawn both individual users and major corporations to Jev is TypeSafe's assertion that it operates much faster and consumes far fewer tokens than LLMs. The company markets its technology as specifically designed for automating tasks, not for producing text or code.

TypeSafe was co-founded in 2024 by Almeida, formerly a researcher at OpenAI, alongside ex-Meta research engineer Sasha Sheng and Erik Gafni, an engineer and entrepreneur.