TypeSafe AI secured $870 million in a Series A funding round valuing the startup at $7.5 billion, led by Andreessen Horowitz. The capital injection comes less than a month after the company launched Jev, a non-text artificial intelligence model designed to automate enterprise tasks by outputting calibrated decisions.
Andreessen Horowitz Backs Series A Valuation
TypeSafe AI closed an $870 million financing round backed by Andreessen Horowitz, alongside participation from Sequoia Capital, existing investor DCVC, and unnamed angel investors. Wilson Sonsini Goodrich & Rosati advised TypeSafe on the deal, with a legal team led by Rob Broderick alongside Estefano Martinez, Hayley Donnell, and Ayush Sahni, Wilson Sonsini announced.
Martin Casado is joining the company’s board. The round terms and investments were confirmed in a post dated October 9, 2026.
The rapid financing follows a surge of attention after the September 15 public debut of Jev. Co-founder Diogo Almeida noted that the viral launch video generated intense inbound interest from chief executives, developers, and investors alike. Prior to the Series A, TypeSafe had emerged from stealth with $40 million in seed financing at a $200 million valuation, reported 36KR.
Jev Trades Text Generation for Typed Decision-Making
Unlike traditional large language models that generate prose token by token, Jev is built on a transformer architecture designed to bypass text output entirely. Instead of writing paragraphs or code, the model produces probabilities and typed values that software applications can process directly.
We have been super good at human language for four years, but it’s not useful for automation because computers speak a different language,
told TechCrunch last month.
Sasha Sheng, a former Meta research engineer, and Erik Gafni, an engineer and entrepreneur, established TypeSafe in 2024 alongside Almeida, an ex-researcher at OpenAI.
When Jev evaluates multiple questions concurrently against a given state, it returns numerical confidence values alongside its answers. Applications use those confidence metrics to catch potential errors before code executes.
Enterprise Adoption Shapes Benchmark Results
TypeSafe reports that about a third of Fortune 500 companies have integrated Jev into their workflows, with the user base surpassing one million users within days of release.
Independent evaluations reveal nuanced performance margins. Testing by researchers on classification tasks showed that Jev offers remarkable speed and cost efficiency compared to various alternatives.
OpenAI Enters the Decision Space
The viral success of non-text classification models has already drawn a corporate response from industry giants. On Oct. 6, just three weeks after Jev’s launch, OpenAI rolled out a competing product, called Decisions API.
TypeSafe plans to allocate its fresh capital toward building out additional machine-native models within its System One lineup, expanding infrastructure, and shipping requested enterprise features as competition across the automated decision market intensifies.
The company stated that its roadmap includes expanding Jev’s functionalities, releasing more machine-native models, delivering infrastructure for intelligent software creation, and rolling out enterprise capabilities demanded by clients.