Magistral Small vs Mistral Large 3
Change modelsMagistral SmallMistral
Mistral Large 3Mistral
IndexAll categories, out of 100.
31.4 ±4.3
34.3 ±4.4 best
GapIndex points below the top.
−2.9 · no clear gap
Top
CoverageIndex weight that has results.
95%
95%
SpeedOutput tokens per second.
—
61/s
EloLMArena rating and rank.
—
1429
ReleasedMonth of release.
2025-06
2025-12
Input / 1MUS dollars per 1M input tokens.
—
$0.50
Output / 1MUS dollars per 1M output tokens.
—
$1.50
1K chats2K tokens in, 500 out, each.
—
$1.75
1K documents20K tokens in, 1K out, each.
—
$11.50
1K agent tasks200K tokens in, 20K out, each.
—
$130.00
ContextMaximum tokens in one request.
128K
262K best
AgenticMulti-step tasks with tools.
24.8
29.7 best
CodingCode writing and repair.
36.3
38.2 best
ReasoningLogic problems and puzzles.
28.2
34.4 best
MultimodalTasks with images and text.
40.0 best
34.5
KnowledgeFacts and expert knowledge.
31.0
38.9 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
12.2
22.9 best
MathMath problems.
41.3 best
34.3
Benchmarks
19 sharedOnly Magistral Small: 10Only Mistral Large 3: 8
Magistral Small
Mistral Large 3