TypeSafe AI, the startup behind the non-text AI model Jev, has raised $870 million at a $7.5 billion valuation — weeks after the model's public release.
The round was led by Andreessen Horowitz, with participation from Sequoia and existing investor DCVC. The fundraise follows a launch on September 15 that TypeSafe says went viral almost immediately: the company claims a third of Fortune 500 companies are already using the model.
Jev is built on a transformer architecture, but it is not a large language model. It does not emit text; it produces probabilities, or what the company calls "calibrated decisions." TypeSafe's pitch is that this makes the model significantly faster and far cheaper in tokens than an LLM, and better suited to automating tasks than to generating prose or code.
"We have been super good at human language for four years, but it's not useful for automation because computers speak a different language," co-founder Diogo Almeida told TechCrunch last month.
Almeida is a former OpenAI researcher. TypeSafe was co-founded in 2024 with Sasha Sheng, previously a research engineer at Meta, and Erik Gafni, an engineer and entrepreneur.
The company's positioning cuts against the prevailing trend of scaling ever-larger language models. If its claims about speed and token efficiency hold up under independent scrutiny, it would offer enterprises a cheaper substrate for the decision-making layers of automation — pricing claims, routing, classification — without paying for generative text they do not need.
That thesis now has unusually large financial backing behind it for a company only weeks past launch. What it does not yet have, publicly, is independent benchmarking.
