Mistral AI

Magistral Small 2506

Zero-eval

by Mistral AI

+
+
+
+
About

Magistral Small was created as an efficient reasoning-focused variant, designed to bring analytical capabilities to applications with tighter resource constraints. Built to balance problem-solving depth with practical deployment needs, it extends reasoning-enhanced AI to broader use cases.

+
+
+
+
Timeline
AnnouncedJun 10, 2025
ReleasedJun 10, 2025
Knowledge CutoffJun 1, 2025
+
+
+
+
License & Family
License
Apache 2.0
Performance Overview
Performance metrics and category breakdown

Overall Performance

4 benchmarks
Average Score
63.2%
Best Score
70.7%
High Performers (80%+)
0
+
+
+
+
All Benchmark Results for Magistral Small 2506
Complete list of benchmark scores with detailed information
AIME 2024
text
0.71
70.7%
Self-reported
GPQA
text
0.68
68.2%
Self-reported
AIME 2025
text
0.63
62.8%
Self-reported
LiveCodeBench
text
0.51
51.3%
Self-reported
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+