The Model Index
Up to four models

Mistral Medium 3.5 vs Mistral Medium 3.1

On the index (GPT-4 = 100), Mistral Medium 3.5 scores 131 and Mistral Medium 3.1 120. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to Mistral Medium 3.5
Presets
Mistral Medium 3.5
131114–148
WeightsOpen
ReleasedApr 2026
Price in / out$1.5 / $7.5
Context262K
Core tests taken1 of 9
Mistral Medium 3.1
120102–138
WeightsClosed
ReleasedAug 2025
Price in / out$0.44 / $2.2
Context131K
Core tests taken0 of 9

Mistral Medium 3.5 and Mistral Medium 3.1 are too close to call (11 points apart, within the uncertainty), at 3.4× the price.

Test by testthird-party results

Mistral Medium 3.5Mistral Medium 3.1
0%25%50%75%100%REASONING & KNOWLEDGEMMLU ProMMLU ProCODINGWeirdMLWeirdMLAGENTSDTBenchDTBenchLMCALMCAPROFESSIONAL WORKCorpFin v2CorpFin v2MortgageTaxMortgageTaxTaxEval v2TaxEval v2
Every result as a table
TestMistral Medium 3.5Mistral Medium 3.1
Reasoning & knowledge
MMLU Pro75.3%75.3%
Coding
WeirdML43.7%33.1%
Agents
DTBench75.5%67.5%
LMCA26.1%19.1%
Professional work
CorpFin v258.8%50.7%
MortgageTax28.9%36.4%
TaxEval v268.0%70.3%

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG