TypeSafe AI raises $870 million, pushing Jev's valuation to $7.5 billion

TypeSafe AI has raised $870 million at a $7.5 billion valuation, just weeks after launching Jev, an AI model that does not generate text. Andreessen Horowitz led the round, with Sequoia and existing investor DCVC also putting in money.

The size of the raise follows the speed of Jev’s takeoff. TypeSafe released the model on September 15, and it gained attention almost immediately. The startup says a third of Fortune 500 companies are already using it, a pace of enterprise adoption the company describes as unusually fast.

Jev is built on a transformer architecture, the same family of design behind most large language models (LLMs), the AI systems that generate text and code. But TypeSafe says Jev is not an LLM. Instead of producing words, it outputs probabilities, which the company calls “calibrated decisions.” TypeSafe claims Jev runs significantly faster and uses far fewer tokens, the chunks of text or data a model processes per request, than comparable language models. The company positions Jev as a tool for automating tasks rather than writing text or generating software.

“We have been super good at human language for four years, but it’s not useful for automation because computers speak a different language,” TypeSafe co-founder Diogo Almeida told TechCrunch last month. Almeida previously worked as a researcher at OpenAI.

Almeida co-founded TypeSafe in 2024 alongside Sasha Sheng, a former Meta research engineer, and Erik Gafni, an engineer and entrepreneur. The company has not detailed Jev’s underlying technical approach beyond saying it outputs probabilities rather than text, and TechCrunch’s reporting does not include independent verification of the claimed speed, token efficiency, or enterprise adoption figures.

What happens next will depend on whether Jev’s performance claims hold up outside TypeSafe’s own telling, and on how a16z, Sequoia, and DCVC expect the company to use $870 million so soon after launch. Benchmarks comparing Jev against established language models, and further detail on which Fortune 500 companies have adopted it, would help clarify how much of the startup’s rapid rise reflects genuine technical advantage versus a well-timed pitch to investors.