Magistral Medium 1.2

Magistral Medium 1.2

Mistral · Released Sep 18, 2025 Specialist
Intelligence #131 / 525
51.5 Our Score
Speed #130 / 244
89.7 tokens / sec
Input #439 / 525
$2.00 per 1M tokens
Output #425 / 525
$5.00 per 1M tokens
Context
Not reported

Analysis Summary

Magistral Medium 1.2 sits in the Specialist tier on our leaderboard, ranked #131 of 525 published models on overall intelligence. At $2.00 input and $5.00 output per 1M tokens, it is among the most expensive on the market.

Editorial notes

Magistral Medium 1.2 from Mistral combines strong math and livecodebench scores with reasonable instruction following, offering a balanced mid-tier option for structured content at $2 input.

Assessed April 26, 2026

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

Performance Profile

Intelligence5.2Technical4.2Value6Content5.5
Intelligence 5.2/10
Technical 4.2/10
Content 5.5/10
Value 6/10

Performance Indices

Source: Artificial Analysis

27.1 Intelligence Index
21.7 Coding Index
32.5 Agentic Index
82 Math Index

Benchmark Scores

Intelligence

GPQA Diamond 73.9% Graduate-level scientific reasoning
HLE 9.6% Humanity's Last Exam
MMLU Pro 81.5% Multi-task language understanding
AIME 2025 82% Competition mathematics (2025)
SciCode 39.2% Scientific computing

Technical

LiveCodeBench 75% Live coding evaluation
TerminalBench Hard 12.9% Agentic terminal tasks
τ²-Bench 52% Conversational agent benchmark

Content

IFBench 43% Instruction following
LCR 51.3% Long-context reasoning

Benchmark data from Artificial Analysis and Hugging Face

How does Magistral Medium 1.2 stack up?

Compare side-by-side with other specialist models.

Compare Models

Model Information

ProviderMistral
Release Date September 18, 2025
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $2.00 $0.002000
Output $5.00 $0.005000

Leaderboard Categories

SEO