TypeSafe AI, the maker of the Jev model, says it has raised a Series A of $870 million at a $7.5B valuation, led by Andreessen Horowitz. The company announced the round in a post on its own blog, read by us on 9 October 2026.
According to the post, Sequoia Capital, existing investor DCVC and a group of angel investors also took part, and Martin Casado is joining TypeSafe’s board. The post says a third of the Fortune 500 are using Jev.
Business Pill · STRUCTURED OUTPUT
A short explainer of structured output: getting a model to answer in a fixed, machine-readable form, the kind of output TypeSafe says Jev is built for. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, the round is a bet on a model that gives up text. TypeSafe’s launch post says Jev returns type-safe structured values with calibrated probabilities instead of strings, and its funding post says a third of the Fortune 500 are using it.
The post is short. It gives the round size and valuation in a footnote: “$870 million at a $7.5B valuation, with Martin Casado joining the board.”
TypeSafe frames the money by audience. For developers, it says there will be more machine-native models. For businesses, it says it has saved customers millions of dollars in production already and will add the enterprise features customers have asked for.
For recruits, the post says: “We are a real company and we’re here to stay.”
Round size and valuation as stated in the footnote of TypeSafe’s funding post: $870 million at a $7.5B valuation, with Martin Casado joining the board. The post’s body names Andreessen Horowitz as the lead.
What Jev Is, in TypeSafe’s Words
TypeSafe’s launch post, written by founder Diogo Almeida, calls Jev its first System One Model: a model built to make fast, structured decisions that software can use directly. Almeida writes that at OpenAI he helped build methods that made language models useful at following instructions, work he says ended up as the research behind ChatGPT.
The launch post says Jev gives up string generation and returns type-safe structured values with calibrated probabilities. It describes a training method TypeSafe calls Reinforcement Learning for Calibrated Decisions (RLCD).
On price, the launch post lists input tokens at $0.042 per million, with output tokens free, against $0.20 to $10 per million input tokens for existing LLMs. It gives Jev’s end-to-end response time as 70ms to 500ms.
TypeSafe’s Own Caveats
The launch post says its claims of 193.6x faster and 444.6x cheaper come from its workflow evals, and that TypeSafe expects these to be on the higher end of real-world gains. It says those reference answers average GPT-6 Astra and Fable 5.1.
On cost, the post says: “We can’t prove it isn’t subsidized”. It adds that TypeSafe expects its pricing to go down, not up.
Posted prices per million input tokens: Jev, from TypeSafe’s launch post, and Microsoft-Decision-1, from Microsoft’s post of 9 October 2026; output tokens are free in both. The LLM range of $0.20 to $10, the 70ms to 500ms response time and the subsidy line are from TypeSafe’s launch post.
The Same Price on Microsoft’s Card
On 9 October 2026, Microsoft introduced Microsoft-Decision-1, a decision-scoring model, at $0.042 per million input tokens with output tokens free. Those are the same figures TypeSafe’s launch post gives for Jev.
Microsoft says its model is built for routing, classification, prioritization, verification and workflow control, and returns a calibrated probability for each answer option. We have no information on how either company set its price.
The Structural Read
The money follows Jev’s early-access launch. TypeSafe’s funding post says it will add the enterprise features customers have asked for and build more machine-native models.
The price is the headline TypeSafe set for itself. Its launch post lists $0.042 per million input tokens with output free, and says TypeSafe expects that pricing to go down, not up.
The performance evidence is TypeSafe’s own. The launch post says its 193.6x and 444.6x figures come from its workflow evals and that TypeSafe expects them to be on the higher end of real-world gains.
TypeSafe AI, funding post
“We find fundraising announcements incredibly boring, so we want to be straightforward with what this means for all of you”
Three Implications
A POSTED PRICE IS NOW SHAREDMicrosoft-Decision-1, launched on 9 October 2026, carries the same $0.042 per million input tokens with output free that TypeSafe lists for Jev.
BOARD SEAT The funding post says Martin Casado is joining TypeSafe’s board.
ENTERPRISE ROADMAP TypeSafe says it has saved customers millions of dollars in production and will add enterprise features.
The framework’s starting point: “Inference is a revenue engine: continuous, distributed across millions of applications and agents, incurring marginal cost with every query.”
As we read it, TypeSafe names its bet on that marginal cost directly: its launch post says Jev is named after William Stanley Jevons, and that every order of magnitude drop in the cost of intelligence unlocks orders of magnitude more use cases.
What Is Not Established
We read TypeSafe’s funding post in full and its Jev launch post except the FAQ answers after the first, which are collapsed on the page and which we did not read. The funding post does not say how much each investor put in, and it does not name the Fortune 500 users. The performance and cost figures are TypeSafe’s own. We did not read press reports of the round, and we did not contact TypeSafe or Andreessen Horowitz.
Business Engineer Framework
The State of The Inference Economy
A Business Engineer framework on how inference, not training, now decides who makes money in AI.
TypeSafe says it raised $870 million at a $7.5B valuation in a Series A led by Andreessen Horowitz, with Sequoia Capital and DCVC taking part and Martin Casado joining the board. Its Jev model is priced at $0.042 per million input tokens with output free, the same posted price as Microsoft-Decision-1.
A note on sourcing. We read TypeSafe’s funding post in full, its Jev launch post except the collapsed FAQ answers after the first, and Microsoft’s Decision-1 post of 9 October 2026; the performance and cost figures are the companies’ own. We did not contact TypeSafe or Andreessen Horowitz. Nothing here is a forecast, and nothing here is financial or investment advice.
Gennaro Cuofano is a business analyst with an International MBA in corporate finance and business strategy (2012). He leads research at FourWeekMBA, which he created, and at The Business Engineer (businessengineer.ai), his newsletter on AI and business strategy, with over 90,000 subscribers and more than 660 published analyses. His work breaks down how companies make money, read from their own filings: revenue and profit structure, capital allocation and ownership. FourWeekMBA reached about four million business people, including C-level executives, investors, analysts and product managers, in 2022 alone. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
