Jev creator TypeSafe closes $870M round at $7.5B valuation

TypeSafe Inc. just pulled in a huge check: $870 million at a $7.5 billion valuation. Andreessen Horowitz led the round, with Sequoia Capital, DCVC and unnamed angel investors also taking part. The timing is striking too. The money arrived less than a month after TypeSafe launched Jev, its artificial intelligence model.

What seems to be driving the enthusiasm is not just the size of the round, but what Jev actually does. The model doesn’t spit out long blocks of text and leave developers to clean them up. It returns structured output from the start, which means applications can use it without as much translation work in the middle. That simpler pipeline can shorten development time and reduce the chances of bugs creeping in.

Jev only handles three kinds of requests: yes-or-no answers, picking from a list, and generating a score. That score can be customized, so one company might use it for cybersecurity severity while another uses it for support-ticket urgency. The model also returns a confidence number when it makes a choice or score, which gives applications a way to account for possible hallucinations.

TypeSafe says Jev was trained with a new approach it calls reinforcement learning for calibrated decisions, along with a new model architecture. The company claims those choices let Jev answer in under 700 milliseconds, making it up to 200 times faster than some frontier LLMs and up to 100 times more cost-efficient. That’s a bold pitch, but the market seems to be listening: TypeSafe says about a third of the Fortune 500 has already adopted the model.

The round is also fuel for a broader plan. Jev is part of a planned family of models called System One, and TypeSafe says the new cash will go toward adding more models and rolling out unspecified enterprise features. In other words, this is less a victory lap than a very expensive starting pistol.