The Pulse

Mistral Acquires SkipLabs as Its Team Forms a New Language Team

Mistral has acquired SkipLabs and says its team will form a new Language team. The group will work on tools to help coding models write and verify software.

AI.info Team · Written with AI · fact-checked · 2026-11-07

Mistral has acquired SkipLabs. The SkipLabs team will form Mistral's new Language team and build tools to help coding models write better code, faster. Models need tools fast enough to run inside their loop, SkipLabs says.

SkipLabs brings deep language and compiler expertise

The SkipLabs team brings expertise in language design, compilers, type systems, static analysis and incremental computation. Its experience draws on decades of building programming languages and developer tools used by thousands of engineers. The team will work at the intersection of programming-language research and frontier AI.

Tools that make code trustworthy were built for humans

The tools that make code trustworthy were built for humans, with slow feedback loops, error messages people must interpret and verification that runs after the fact, SkipLabs says. Models need feedback they can learn from and languages they can reason in, the company says. Mistral's post says: “The tools that make code trustworthy were built for humans. Models need the opposite. Let's reinvent them from scratch.”

Team will shape every stage of the model, from training and evaluation to how agents write and verify software

The team will explore how languages, type systems, compilers and program analysis can shape the next generation of models and coding agents. The team will build tools to make coding models write better code, faster.

AI breaks tools' assumptions, Julien Verlaguet says

“I've spent my career building languages and type systems designed to catch mistakes before they shipped. Every tool we ever built assumed there was a human on the other end, someone reading the error message, judging what mattered, choosing what to fix. With AI, the author is no longer a person. That breaks many of the assumptions these tools rest on and that's precisely the exciting part: we get to reinvent them from scratch. At Mistral, we get to push these ideas all the way down into how models are trained.”

Julien Verlaguet, founder of SkipLabs