Technology

TypeSafe AI raises US$870 million weeks after launching its Jev model

TypeSafe AI closes a US$870 million round at a US$7.5 billion valuation, weeks after launching Jev. What it means for the cost of your operation.

TypeSafe AI, the company behind the Jev model, closed a US$870 million round at a US$7.5 billion valuation, led by the fund Andreessen Horowitz and announced on October 9, 2026. The deal comes less than a month after the launch of Jev, presented on September 15.

According to TechCrunch, Sequoia Capital and previous investor DCVC took part in the round. Jev is not a language model: it uses a transformer architecture, but instead of writing text it returns probabilities that the company calls calibrated decisions, designed so that software can act without human intervention. SiliconANGLE reported that the company claims Jev responds in under 700 milliseconds and is up to 200 times faster and up to 100 times cheaper than frontier models. The same outlet noted that the funds will go toward expanding the System One model series and features for large enterprises. For an operator in Mexico or Colombia, the difference is one of cost: automating a decision is billed per token processed, not per conversation.

The model was introduced on September 15, 2026 and accepts three types of request: a yes-or-no answer, selection of one item from a list, and assignment of a score, according to SiliconANGLE. When it returns a score or a selection, it also delivers a confidence number that the application uses to limit the effect of a wrong answer. TypeSafe maintains that one third of Fortune 500 companies already use Jev, without identifying which ones, and Infobae noted that this validation currently rests on the company's own reports. The company was founded in 2024 and its founders are Diogo Almeida, a former OpenAI researcher; Sasha Sheng, a former research engineer at Meta; and Erik Gafni, an engineer and entrepreneur.

The news leaves a cost distinction on the table: which API calls need to generate language and which only need a decision. The indicator to watch is the cost per decision and how dependent the product becomes on a provider that had been on the market for only weeks.

This piece was written with the assistance of artificial intelligence from verified sources and reviewed by a human editor before publication.

Sources