Magistral Small 1
Magistral Small 1Rankingin thinking mode?
Alternatives?
CheaperMiMo v2.6 ProthinkingOpen×5.6-33%$0.13Memory (4-bit): 14 GB → 615 GB+4
BetterGPT-6 Lunalow×2.6-75%$0.005Function calling only with none · Chat Completions+3
OpenQwen3.8 27Bxhigh×4.1+17%$1.01Memory (4-bit): 14 GB → 17 GB+5
CheaperDeepSeek V4.1 Flash 1017maxOpen×5.5-36%$0.26Memory (4-bit): 14 GB → 458 GB+5
CheaperClaude Haiku 5.5max×5.3-75%$0.21Remove sampling parameters+5MiMo v2.6 ProthinkingOpen×5.6-33%$0.13Memory (4-bit): 14 GB → 615 GB+4DeepSeek V4.1 Flash 1017maxOpen×5.5-36%$0.26Memory (4-bit): 14 GB → 458 GB+5Claude Haiku 5.5max×5.3-75%$0.21Remove sampling parameters+5
MiniMax M3.1 Flashthinking×5.2-68%$0.22Complete model response must be appended
GLM 5.3 FlashmaxOpen×5.1-71%$0.28Memory (4-bit): 14 GB → 193 GB+5GPT-6 Lunalow×2.6-75%$0.005Function calling only with none · Chat Completions+3GPT-5.6 Lunalow×2.6-45%$0.01Mistral Small 4thinkingOpen+37%-68%$0.015Memory (4-bit): 14 GB → 72 GBMiMo v2.5thinkingOpen×2.7-79%$0.019Memory (4-bit): 14 GB → 186 GB
Granite 4.2 8BthinkingOpen+35%-87%$0.024Qwen3.8 27BxhighOpen×4.1+17%$1.01Memory (4-bit): 14 GB → 17 GB+5IK2 Horizon MoVA 36B A4BthinkingOpen×3.1+115%$1 / $4Memory (4-bit): 14 GB → 22 GBQwen 3.5 27BthinkingOpen×2.8same$0.30 / $2.40Memory (4-bit): 14 GB → 17 GBAG9v3 39A5BOpen×2.7——Memory (4-bit): 14 GB → 23 GBQwen3.6 27BthinkingOpen×2.6+66%$0.62Memory (4-bit): 14 GB → 17 GBMistral Large 4thinking×4.7—$1.13Infinite whitespace stream+4Mistral Medium 3.5highOpen+73%—$0.50Memory (4-bit): 14 GB → 77 GBMistral Small 4thinkingOpen+37%—$0.015Memory (4-bit): 14 GB → 72 GBMistral Large 3non-thinkingOpen+13%—$0.031Memory (4-bit): not verifiedMagistral Medium 1thinkingsame——Context: 128K → 40KDevstral Mediumnon-thinkingOpensame—$0.40 / $2Memory (4-bit): not verifiedMistral Small 3non-thinkingOpen-19%—$0.05 / $0.08Context: 128K → 32KMinistral 3 14B 2512non-thinkingOpen-27%—$0.019All Mistral models →Run it yourself
estimateHistory
2 versionsBenchmarks
all categoriesFAQ
How much does Magistral Small 1 cost?›
$0.70 per 1M input tokens and $1.16 per 1M output tokens (cheapest host: Featherless).
Is Magistral Small 1 free to use?›
There is no free API; the cheapest paid price is $0.70 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Magistral Small 1, and what is it best at?›
#213 of 334 in Chat & Knowledge, #161 of 239 in Coding, #161 of 242 in Agents & Tool Use, #230 of 245 in Writing & Language.
Is Magistral Small 1 better than Qwen3 VL 32B Instruct?›
On AA Intelligence Index 4.3.2, Magistral Small 1 scores 8.2 vs 8.4 — about the same.
How can I use Magistral Small 1?›
Through the API of Featherless.
What is the context window of Magistral Small 1?›
128K tokens.
When was Magistral Small 1 released? Is there a newer version?›
Released 2025-06-10. It has been superseded by Magistral Small 1.2.
Is Magistral Small 1 open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Magistral Small 1 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Magistral Small 1 read images?›
No, it accepts text only.
What is Magistral Small 1?›
Magistral Small 1 is an open-weights language model from Mistral, released Jun 2025.
What are the best alternatives to Magistral Small 1?›
MiMo v2.6 Pro (cheaper), GPT-6 Luna (better), Qwen3.8 27B (open weights).