FrontierThe story, in brief
Mistral AI Launches Remote Agents in Vibe and Mistral Medium 3.5 with 77.6% SWE-Bench Verified Score
77.6% on SWE-Bench Verified. Mistral's new 128B model just closed the gap with closed-source leaders — and shipped agentic coding agents to production.
The KeyNews take
Why it matters
Mistral AI is moving from model-first to agent-first positioning with a capability benchmark score that legitimizes its flagship against GPT/Claude, while shipping product (remote agents in Le Chat) that developers can use immediately. This signals Mistral's bet on agents-as-capability as a differentiation vector.
The key facts
5 to knowMistral Medium 3.5: 128B flagship model
77.6% SWE-Bench Verified score
Remote agents feature in Vibe (async cloud-based coding sessions)
Agentic Work mode in Le Chat
Published May 2, 2026
Go to the source
MarkTechPostmarktechpost.com
Publisher excerpt: Mistral AI's latest release brings async cloud-based coding sessions, a new 128B flagship model, and an agentic Work mode to Le Chat — a meaningful step forward for developers building with AI agents.