Mistral Small 3
Mistral Small 3Alternatives?
CheaperGPT OSS 20BhighOpen×2.9-37%$0.011BetterGPT-6 Lunalow×5.4+248%$0.005Function calling only with none · Chat Completions+3
OpenQwen3.8 27Bxhigh×9+1,563%$1.01Memory (4-bit): 14 GB → 17 GB+5
CheaperNemotron 3.5 LightningthinkingOpen×2.8-15%$0.079Memory (4-bit): 14 GB → 20 GB
CheaperGranite 4.2 3BthinkingOpen+76%-9%$0.006GPT OSS 20BhighOpen×2.9-37%$0.011Nemotron 3.5 LightningthinkingOpen×2.8-15%$0.079Memory (4-bit): 14 GB → 20 GBGranite 4.2 3BthinkingOpen+76%-9%$0.006Qwen3 0.6BthinkingOpen+50%-13%$0.04 / $0.08Qwen3.5 0.8Bnon-thinkingOpen+36%-65%$0.01 / $0.05GPT-6 Lunalow×5.4+248%$0.005Function calling only with none · Chat Completions+3Granite 4.2 3BthinkingOpen+76%-9%$0.006GPT-5.6 Lunalow×5.3+683%$0.01GPT OSS 20BhighOpen×2.9-37%$0.011Mistral Small 4thinkingOpen×2.6+357%$0.015Memory (4-bit): 14 GB → 72 GBQwen3.8 27BxhighOpen×9+1,563%$1.01Memory (4-bit): 14 GB → 17 GB+5Qwen 3.5 27BthinkingOpen×6.4+1,335%$0.30 / $2.40Memory (4-bit): 14 GB → 17 GB
Gemma 4 31BthinkingOpen×6.3+178%$0.10 / $0.34Memory (4-bit): 14 GB → 19 GBIK2 Horizon MoVA 36B A4BthinkingOpen×6.2+2,943%$1 / $4Memory (4-bit): 14 GB → 22 GBQwen3.6 27BthinkingOpen×6.1+2,248%$0.62Memory (4-bit): 14 GB → 17 GBMistral Large 4thinking×9.3—$1.13Infinite whitespace stream+4Mistral Medium 3.5highOpen×3.7—$0.50Memory (4-bit): 14 GB → 77 GBMistral Small 4thinkingOpen×2.6—$0.015Memory (4-bit): 14 GB → 72 GBMagistral Medium 1thinking×2.6——Magistral Small 1thinkingOpen×2—$0.70 / $1.16Ministral 3 3B 2512Open+43%—$0.008Ministral 3 14B 2512non-thinkingOpensame—$0.019Ministral 3 8B 2512Opensame—$0.011All Mistral models →Run it yourself
estimateHistory
3 versionsBenchmarks
in Chat & KnowledgeFAQ
How much does Mistral Small 3 cost?›
$0.05 per 1M input tokens and $0.08 per 1M output tokens (cheapest host: DeepInfra).
Is Mistral Small 3 free to use?›
There is no free API; the cheapest paid price is $0.05 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Mistral Small 3, and what is it best at?›
#316 of 334 in Chat & Knowledge, #129 of 152 in Coding, #194 of 242 in Agents & Tool Use, #168 of 195 in Writing & Language.
Is Mistral Small 3 better than Phi 4?›
On Humanity's Last Exam, Mistral Small 3 scores 3.8 vs 3.8 — about the same.
How can I use Mistral Small 3?›
Through the API of DeepInfra, Featherless.
What is the context window of Mistral Small 3?›
32K tokens, with up to 16K output tokens.
When was Mistral Small 3 released? Is there a newer version?›
Released 2025-01-30. It is the current version in its line.
Is Mistral Small 3 open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Mistral Small 3 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Mistral Small 3 read images?›
No, it accepts text only.
What is Mistral Small 3?›
Mistral Small 3 is an open-weights language model from Mistral, released Jan 2025.
What are the best alternatives to Mistral Small 3?›
GPT OSS 20B (cheaper), GPT-6 Luna (better), Qwen3.8 27B (open weights).