TypeSafe raises $870M at $7.5B on four different Fortune 500 figures

TypeSafe AI raised $870 million at a $7.5 billion valuation on 9 October, and four announcements published that same day gave four different figures for how much of the Fortune 500 uses its model. The lowest, 25%, sits in the post from Andreessen Horowitz, the firm that led the round. The company’s own blog said a third. But no write-up we’ve read set the figures side by side.

The round was led by a16z with participation from Sequoia Capital, existing investor DCVC and a group of angel investors, according to TypeSafe’s announcement. That post carries two things which the trade write-ups left out. It calls the money a Series A, and it says Martin Casado is joining the board. Neither TechCrunch nor SiliconANGLE named the stage or the board seat.

Four numbers for one adoption claim

Casado is one of the authors of the a16z post, which is where the smallest figure appears: “25% of the fortune 500 enterprises have integrated Jev.” Diogo Almeida, TypeSafe’s chief executive, posted 29.4% from his own account the same afternoon, in a message quoted in full by Latent Space. Bloomberg’s write-up, summarised on Techmeme, says about 33%, citing the company.

Source, all 9 October 2026	Fortune 500 share stated	Companies implied
a16z, the lead investor	25%	125
Diogo Almeida on X	29.4%	147
Bloomberg, citing the company	about 33%	about 165
TypeSafe blog, TechCrunch, SiliconANGLE	a third	about 167

So the claim runs from 125 companies to roughly 167, a spread of about 42 names by our arithmetic. That matters because adoption is doing most of the work in this story. Jev only shipped on 15 September, so enterprise uptake is the main evidence that $7.5 billion attaches to a business rather than to a launch video. A quarter of the Fortune 500 and a third of it are not the same claim.

There’s also a figure none of the briefed coverage carried at all. Sequoia partner Pat Grady said Jev went from zero to $100 million in revenue in its first seven days, in a video on the state of AI posted on 27 September, which the launch agency Doomers cites in its case study. That same case study notes TypeSafe hasn’t published a revenue figure of its own. Latent Space rendered the number as more than $100 million in annual recurring revenue, which isn’t the same measure as revenue booked in a week.

What an outside source can actually see

One independent reading does exist, and it’s narrower than the Fortune 500 claim. Vercel reported on 18 September that nearly 13% of its paid teams were calling Jev within 24 hours of it reaching AI Gateway, twice the share the GPT-5.6 family reached and more than six times Fable 5.1’s. Vercel called it the fastest-adopted model in the gateway’s history. Those are Vercel’s own customers, not the Fortune 500, so it counts something smaller and checkable.

TypeSafe left stealth on 15 September with $40 million led by DCVC, and Forbes called it a $200 million startup that day, as we reported when Jev first shipped. Reaching $7.5 billion 24 days later is a 37.5 times increase on that figure, by our arithmetic. a16z says Jev hit a trillion generated tokens three days after launch, and puts it at roughly a hundredth to a five-hundredth of the cost of frontier models.

Call it the Jev-ons paradox: make intelligence cheap enough to call anywhere, and it gets called everywhere.

a16z, Investing in TypeSafe AI

The performance figures wobble in the same direction. SiliconANGLE reported that Jev answers in under 700 milliseconds, making it up to 200 times faster than some frontier language models and up to 100 times more cost-efficient. TypeSafe’s own launch post gives end-to-end response times of 70 to 500 milliseconds, and traces its homepage claims of 193.6 times faster and 444.6 times cheaper to four published workflow evals. So on cost, the trade write-up is the conservative one by a wide margin.

What’s worth watching is whether anyone outside the company ever pins the adoption number down. Enterprise logos are the easiest thing to assert and the hardest to verify, and they’re usually among the parts of a round announcement nobody confirms. Almeida told TechCrunch in September that “we have been super good at human language for four years, but it’s not useful for automation because computers speak a different language”. Almost every figure above traces back to TypeSafe or to an investor in it, which is why Vercel’s gateway data stands out as the exception.