TypeSafe AI Raises $870 Million for Jev as OpenAI, Amazon and Cloudflare Ship Rivals

TypeSafe AI announced an $870 million Series A on Friday to expand Jev, its decision model for software. The Andreessen Horowitz-led financing values the company at $7.5 billion. The funding follows Jev's launch last month as OpenAI, Amazon Web Services and Cloudflare release competing decision models.

Funding and plans

The October 9 round included Sequoia Capital, existing investor DCVC and angel investors. Andreessen Horowitz general partner Martin Casado is joining the board. The valuation came 24 days after Jev's September 15 debut, when TypeSafe emerged with a $40 million seed round led by DCVC.

Diogo Almeida, Erik Gafni and Sasha Sheng founded TypeSafe in 2024. Almeida, its CEO, previously helped develop reinforcement learning from human feedback at OpenAI and worked on InstructGPT, ChatGPT and GPT-4.

Jev returns typed choices, scores and yes/no probabilities for software to consume instead of text. Developers define the allowed answers and confidence thresholds: a customer-service program could route messages to billing or technical support automatically, sending low-confidence messages to a person. Jev does not supply a written explanation. As of October 9, TypeSafe claimed Jev processed requests in under 700 milliseconds, up to 200 times faster than some frontier LLMs.

It is the first of TypeSafe's planned System One models. The company said it plans more machine-native models and infrastructure for smart software, alongside enterprise features requested by customers.

Conflicting adoption claims

TypeSafe's October 9 post said a third of Fortune 500 companies, about 167 out of 500, were using Jev. Its lead investor's post, published the same day, put the share at 25%, about 125 out of 500, and described those companies as having integrated the model. Neither account explained the difference.

TypeSafe also said that it had saved production customers millions of dollars. Its post names no customer and gives no figure beyond “millions.”

Nearly 13% of Vercel's AI Gateway paid teams tried Jev during its first 24 hours there. Vercel said that was more than twice the paid-team reach of any previous model launch.

Competing decision models

OpenAI released its Decisions API on October 6, using GPT-6 Luna without additional training for the initial version. The service accepts images as well as text; Jev accepts text only. As of October 8, OpenAI listed input pricing at $0.10 per million tokens, more than twice TypeSafe's $0.042 per million for Jev. Neither charged for output tokens.

OpenAI employee Nikunj Handa credited Jev with “inspiring this whole thing” and said the Decisions API had not been on the roadmap four weeks earlier.

AWS released the open-source Strands Decider 2B on October 1 under the Apache 2.0 license. Developers supply the hardware or cloud resources to run it. AWS's benchmark chart published with the release used public JevBench results but left out Jev and the checkpoint AWS planned to release.

By October 9, Cloudflare had introduced Clef and Clef-flash through Workers AI, offering decision models that enterprises could run closer to their applications.

Evidence and operating costs

TypeSafe's speed, cost, savings and enterprise adoption figures come from the company and its investors. Stanford and Nvidia researchers found their open CLM-8B model up to nine times faster than Jev on some tasks but less accurate at tool-calling, in tests specific to their workloads and setup. TypeSafe has not released Jev's weights or full training recipe.

“A decision model might reduce inference spending significantly, but if the enterprise needs additional engineering resources to maintain multiple models, manage failures, and investigate inconsistent results, the financial benefit may be smaller than expected,” said Aditya Ranjan, a senior data engineer at supermarket chain H-E-B.

Hang Ten Systems Raises $53 Million Five Weeks After Its First Seed Round

Hang Ten Systems raised an additional $53 million in a round led by Xora, a Temasek-backed fund, after former Infosys CEO Vishal Sikka founded the enterprise AI services company in May 2026. The two s

The Implicator

Naive AI Hits $1.4 Billion Valuation After Raising $400 Million in Seven Months

Naive AI has raised $400 million at a $1.42 billion valuation. The figures come from one unnamed person with direct knowledge cited in The Information’s exclusive, and Naive AI has not confirmed the f

The Implicator

OpenAI and Anthropic Drive California Venture Funding to $366 Billion

California-based companies have announced $366 billion in venture funding since the start of 2026, a state record as of Aug. 20. OpenAI and Anthropic supplied more than half of that total through a ha

The Implicator

Marcus Schuler

San Francisco

Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai