DEV Community

Cover image for TypeSafe raises $870M Series A at $7.5B for Jev
techaiwire
techaiwire

Posted on Originally published at techaiwire.com

TypeSafe raises $870M Series A at $7.5B for Jev

TypeSafe AI, the maker of the Jev model, raised $870 million at a $7.5 billion valuation in a round announced on October 9, 2026. Andreessen Horowitz led it. The round came 24 days after TypeSafe left stealth with a $40 million seed round, valued then at about $200 million. Few startups have repriced that fast, and the money now has to turn early interest into dependable revenue.

The round

Unite.AI reports the round as a Series A. Sequoia Capital took part, alongside existing investor DCVC and a group of angel investors. Martin Casado, a general partner at Andreessen Horowitz, is joining TypeSafe's board.

Detail TypeSafe AI
Amount $870 million
Valuation $7.5 billion
Lead investor Andreessen Horowitz
Other investors Sequoia Capital, DCVC, angel investors
Previous round $40 million seed, led by DCVC, announced September 15, 2026
Seed valuation About $200 million post-money, per Crypto Briefing
Founded 2024, in San Francisco

The reports differ on a few labels. Dealroom calls it a late-stage round rather than a Series A. It ranks the deal in the 97th percentile of all US late-stage deep-tech rounds. RuntimeWire, citing Bloomberg, notes that no closing date was reported.

Crypto Briefing adds that some investors had already discussed a round above $1 billion, at a valuation above $10 billion.

What Jev does

Jev is not a chatbot. It takes a program's current state and returns a typed value, a structured answer that fits a format the developer defines. TypeSafe calls this a System One model. Tech AI Wire covered how Jev returns typed decisions instead of text when it launched in early access on September 15.

TypeSafe prices it at $0.042 per million input tokens, with output free. A token is a small chunk of text. The company claims answers in 70 to 500 milliseconds, end to end.

The team comes from large AI labs. Crypto Briefing names Diogo Almeida as chief executive, Erik Gafni as chief technology officer and Sasha Sheng as chief operating officer. Their past employers include OpenAI, Google Brain and Meta's FAIR research group.

The traction claims

Most of the usage figures come from the company or its lead investor, and they do not all match:

  • Fortune 500 use. TypeSafe says about a third of Fortune 500 companies use Jev. Unite.AI reports that a16z puts it at 25%. Neither names the customers.
  • Users. Dealroom reports the startup's claim that Jev passed one million users.
  • Volume. According to a16z, Jev processed 1 trillion tokens within three days of launch.

One customer case study gives harder numbers. TypeSafe said on October 7 that the hiring company Jack & Jill cut its screening cost by 88%. The cost fell from $0.755 to $0.092 per 1,000 candidates. Median screening time dropped from 20.3 to 10.3 seconds.

Where the money goes

According to Pulse 2.0's report of the announcement, the funds will expand Jev and build more models designed for machines rather than chat. They will also scale TypeSafe's infrastructure and add enterprise features for security, governance and integration.

What this means for developers

For teams weighing Jev, the round changes the risk more than the product.

  • Supply looks safer. A company with $870 million in the bank is less likely to vanish mid-integration. That matters for a model with no direct drop-in replacement.
  • Expect enterprise controls next. Security, governance and integration features are named uses of the money. If your company needs audit logs or single sign-on before approving a tool, watch for those releases.
  • Do not plan around today's price. Free output and $0.042 input are launch terms. Investors paying a $7.5 billion valuation will expect revenue, so budget for change.
  • Verify the case study on your own data. An 88% cost cut on candidate screening is one customer's result. Run Jev against your current model on one real decision before switching.
  • Keep an open fallback in view. Laya is an open-weights model of the same shape that you can run yourself, which limits lock-in.

This article was first published on Tech AI Wire.

Also available in

Deutsch · 日本語 · Français · Español · Português

Related on Tech AI Wire

Sources

Top comments (0)