Jev is TypeSafe AI's flagship System One Model, returning typed probabilistic decisions instead of generated text.
View competitors
TypeSafe AI is a San Francisco AI lab building machine-native models that return typed decisions.
TypeSafe AI is an artificial intelligence research laboratory based in San Francisco, California. The company builds machine-native intelligence infrastructure designed to make decisions inside software rather than generate text for people. Its models return typed outputs carrying calibrated confidence estimates, so applications can act autonomously when confidence is high and escalate when it is not.
The company emerged from stealth with $40 million in seed funding led by DCVC. It was founded by Diogo Almeida, a former OpenAI researcher and co-inventor of reinforcement learning from human feedback, together with Erik Gafni and Sasha Sheng. Its first public model, Jev, is a System One Model trained with Reinforcement Learning for Calibrated Decisions.

Superagent
superagent.sh

Microsoft
microsoft.com

Cloudflare
cloudflare.com

Preference Model

Thinking Machines Lab

Experiential Labs

Anthropic
anthropic.com

River AI
river.ai

Tracer
tracerml.ai

Patronus AI
patronus.ai

Prime Intellect
primeintellect.ai

General Intuition

attimet
attimet.com

Reflection AI
reflection.ai

Fastino
fastino.ai

CalypsoAI
calypsoai.com

JustAI
getjust.ai

Klavis AI
klavis.ai

hiloop
hiloop.ai

Halluminate
halluminate.ai
Companies similar to TypeSafe AI include Superagent, Microsoft, Cloudflare and Preference Model.
TypeSafe AI was founded in 2024.
TypeSafe AI's headquarters is located in San Francisco, CA, US.
TypeSafe AI's most recent funding round was for $870M (USD) in October 2026.
As of Oct 9, 2026, TypeSafe AI has disclosed funding of $910M (USD).
TypeSafe AI has 11-50 employees as of Sep 17, 2026.