Mistral releases open and hosted Magistral reasoning models
Mistral released Magistral Small 2506 as open weights and Magistral Medium 2506 through its API and Le Chat, adding explicit reasoning to its model portfolio.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why it moved the index
Distinct reasoning models increased Mistral's competitive capability and distributed an open-weight tier without provider-side access controls alongside a broadly hosted frontier tier.
Assessment history
-
R1
Toward 48 · confidence 95
New exact reasoning-model family with no durable event or model collision.
12 Aug 2026
Share this page
-
DoomBench assesses “Mistral releases open and hosted Magistral reasoning models” as evidence moving toward doom, with magnitude 48 and confidence 95 out of 100 in the capability gains category.
-
The DoomBench assessment of “Mistral releases open and hosted Magistral reasoning models” is based on reporting from Mistral AI and records the editorial rationale, source quality, attribution, and revision history.
-
DoomBench summarizes “Mistral releases open and hosted Magistral reasoning models” as follows: Mistral released Magistral Small 2506 as open weights and Magistral Medium 2506 through its API and Le Chat, adding explicit reasoning to its...
https://www.doombench.com/news/mistral-releases-open-and-hosted-magistral-reasoning-models-2025-06-10