llm-benchmarks
githubdrose.io

mistral Provider Benchmarks


Measured over the last 30 days

Models Tracked
14
Avg Tokens / Second
48.89
Avg Time to First Token (ms)
1465.00
Last Updated
Oct 17, 2026

All mistral models

14 live · one row per model · fastest first · 1 retired lanes not listed · 2 folded behind their current lane
mistral site
ModelAvg Toks/SecMinMaxAvg TTF (ms)
Ministral 3 3B 251275.5075.5075.50510.00
Mistral Medium 364.6064.6064.60600.00
Voxtral Small 24B 250762.2062.2062.20680.00
Mistral Small 458.5058.5058.50600.00
Mixtral 8x22B Instruct55.8055.8055.80640.00
Ministral 3 14B 251255.0055.0055.00560.00
Codestral 250852.7052.7052.70780.00
Saba51.5051.5051.50760.00
Mistral Medium 3.150.8050.8050.80750.00
Ministral 3 8B 251248.7048.7048.70850.00
Mistral Medium 3.547.7029.4066.30880.00
Mistral Large 3 251233.6033.2034.00830.00
Mistral Large22.0022.0022.001670.00
Mistral Large 45.924.487.2610400.00

Frequently Asked Questions

Which mistral model is fastest?
Based on recent tests, Ministral 3 3B 2512 shows the highest average throughput among tracked mistral models.
How many recent measurements feed this dashboard?
This provider summary aggregates 439 individual prompts measured across 431 monitoring runs over the past month.