Mistral Small 3.1 24B
Mar 2025
Mistral Small 3.1 24BCompare inon
Benchmarks of Mistral Small 3.1 24B20
BenchmarkByScoreRank
Knowledge18
Multimodal1
Writing1
No benchmark matches
Rankingconfiguration unstated?
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperGemma 4 31BunstatedOpen+80%-60%$0.10 / $0.34Memory (4-bit): 14 GB → 19 GB
BetterGPT-6 Astraunstated×2.9+4,875%$10 / $50Monitoring can return HTTP 403+2
OpenMuse Glimmer 30Bunstated×2.4+24%$0.30 / $1.10Memory (4-bit): 14 GB → 18 GBCheaperGemma 4 26B A4BunstatedOpen+72%-35%$0.15 / $0.60Memory (4-bit): 14 GB → 15 GBCheaperGemma 4 12BunstatedOpen+70%-63%$0.10 / $0.30ModelScoreCostPer taskWhat changes
Gemma 4 31BunstatedOpen+80%-60%$0.10 / $0.34Memory (4-bit): 14 GB → 19 GBGemma 4 26B A4BunstatedOpen+72%-35%$0.15 / $0.60Memory (4-bit): 14 GB → 15 GBGemma 4 12BunstatedOpen+70%-63%$0.10 / $0.30Gemma 3 27BunstatedOpen+67%-75%$0.08 / $0.16Memory (4-bit): 14 GB → 16 GB
QwQ 32BOpen+66%-47%$0.15 / $0.40Memory (4-bit): 14 GB → 20 GB+1ModelScoreCostPer taskWhat changes
GPT-6 Astraunstated×2.9+4,875%$10 / $50Monitoring can return HTTP 403+2
Claude Fable 5.1unstated×2.8+4,875%$10 / $50Replace forced tool_choice+5Claude Opus 5unstated×2.8+2,388%$5 / $25GPT-6 Solunstated×2.8+895%$2 / $10
Kimi K3unstatedOpen×2.7+1,393%$3 / $15Memory (4-bit): 14 GB → 1,668 GB+4ModelScoreCostPer taskWhat changes
Muse Glimmer 30BunstatedOpen×2.4+24%$0.30 / $1.10Memory (4-bit): 14 GB → 18 GBQwen3.8 27BunstatedOpen×2.2+138%$0.42 / $2.55Memory (4-bit): 14 GB → 17 GB+5Gemma 4 31BunstatedOpen+80%-60%$0.10 / $0.34Memory (4-bit): 14 GB → 19 GBGemma 4 26B A4BunstatedOpen+72%-35%$0.15 / $0.60Memory (4-bit): 14 GB → 15 GBGemma 4 12BunstatedOpen+70%-63%$0.10 / $0.30ModelScoreCostPer taskWhat changes
Mistral Large 3unstatedOpen+86%—$0.50 / $1.50Memory (4-bit): not verifiedMistral NemounstatedOpen+16%—$0.15 / $0.15No image inputMistral Small 3unstatedOpen-7%—$0.05 / $0.08Context: 128K → 32K+1Mistral Large 4thinkingn/a—$1.61Infinite whitespace stream+4Mistral Medium 3.5highOpenn/a—$0.43Memory (4-bit): 14 GB → 77 GBMistral Small 4thinkingOpenn/a—$0.017Memory (4-bit): 14 GB → 72 GBMistral Smalln/a—$1 / $3Context: 128K → 32KMinistral 3 14B 2512non-thinkingOpenn/a—$0.019All Mistral models →Run it yourself
estimateFits on✗ 8 GB (phone, laptop)✓ 16 GB · compressed✓ 24 GB GPU · compressed✓ 48 GB GPU · 8-bit✓ 80 GB GPU · full quality
LicenceApache-2.0free for commercial use
History
2 versionsBenchmarks
all categoriesChat & Knowledge
Arena Text · overall1,303#154
Arena Text · Chinese1,252.8#161
Arena Text · English1,294.6#166
Arena Text · French1,272.8#139
Arena Text · German1,266.5#129
Arena Text · instruction following1,263.3#158
Arena Text · Japanese1,207.6#119
Arena Text · Korean1,206#127
Arena Text · non-English1,254.6#161
Arena Text · Polish1,241.5#132
Arena Text · Russian1,263.4#158
Arena Text · Spanish1,283.4#140
EQ-Bench 3832.2#48
Arena Text · overall · October 22, 20261,277.1#158
Arena Text · overall · October 15, 20261,277.1#151
Writing & Language
Arena Text · creative writing1,270#148
EQ-Bench Creative Writing v3758.9#88
Vision & Documents
Arena Vision · overall1,136.5#77
FAQ
How much does Mistral Small 3.1 24B cost?›
$0.35 per 1M input tokens and $0.56 per 1M output tokens (cheapest host: Cloudflare).
Is Mistral Small 3.1 24B free to use?›
There is no free API; the cheapest paid price is $0.35 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Mistral Small 3.1 24B, and what is it best at?›
#166 of 205 in Chat & Knowledge, #150 of 197 in Writing & Language, #77 of 88 in Vision & Documents.
Is Mistral Small 3.1 24B better than Llama 4 Scout?›
On EQ-Bench Creative Writing v3, Mistral Small 3.1 24B scores 758.9 vs 781.1 — lower.
How can I use Mistral Small 3.1 24B?›
Through the API of Cloudflare.
What is the context window of Mistral Small 3.1 24B?›
128K tokens, with up to 102K output tokens.
When was Mistral Small 3.1 24B released? Is there a newer version?›
Released 2025-03-11. It is the current version in its line.
Is Mistral Small 3.1 24B open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Mistral Small 3.1 24B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Mistral Small 3.1 24B read images?›
It accepts text and images as input.
What is Mistral Small 3.1 24B?›
Mistral Small 3.1 24B is an open-weights language model from Mistral, released Mar 2025.
What are the best alternatives to Mistral Small 3.1 24B?›
Gemma 4 31B (cheaper), GPT-6 Astra (better), Muse Glimmer 30B (open weights).