Ministral 3 3B 2512
Ministral 3 3B 2512Alternatives?
CheaperLing 3.0 Flash VLthinkingOpen×23.6-10%$0.06 / $0.18Memory (4-bit): 1.9 GB → 75 GB
BetterGPT-6 Lunalow×18.1-42%$0.005Function calling only with none · Chat Completions+3IOpenK2 Horizon 7Bthinking×31.4+425%$0.30 / $1.20Memory (4-bit): 1.9 GB → 4.5 GB+1CheaperLing 3.0 Flash FinthinkingOpen×11.1-10%$0.06 / $0.18Memory (4-bit): 1.9 GB → 76 GB+1CheaperLing 3.0 FlashthinkingOpen×4.8-10%$0.06 / $0.18Memory (4-bit): 1.9 GB → 76 GB+1Ling 3.0 Flash VLthinkingOpen×23.6-10%$0.06 / $0.18Memory (4-bit): 1.9 GB → 75 GBGPT-6 Lunalow×18.1-42%$0.005Function calling only with none · Chat Completions+3Ling 3.0 Flash FinthinkingOpen×11.1-10%$0.06 / $0.18Memory (4-bit): 1.9 GB → 76 GB+1Ling 3.0 FlashthinkingOpen×4.8-10%$0.06 / $0.18Memory (4-bit): 1.9 GB → 76 GB+1
Granite 4.2 3BthinkingOpensame-24%$0.006Context: 131K → 128K+1GPT-6 Lunalow×18.1-42%$0.005Function calling only with none · Chat Completions+3GPT-5.6 Lunalow×17.6+25%$0.01
MiMo v2.5thinkingOpen×27.7+143%$0.019Memory (4-bit): 1.9 GB → 186 GB
Claude Haiku 5.5low×34.4+212%$0.024Remove sampling parameters+5GPT-5 Minihigh×9.7+582%$0.053Granite 4.2 8BthinkingOpen×3.1+201%$0.024Context: 131K → 128K+2OMiniCPM5 2BthinkingOpen×2.7——No image input
Qwen 3.5 9BthinkingOpensame+2,609%$0.21Memory (4-bit): 1.9 GB → 4.8 GBMistral Large 4thinking×90+14,353%$1.13Infinite whitespace stream+4Mistral Medium 3.5highOpen×9.5+6,298%$0.50Memory (4-bit): 1.9 GB → 77 GBMistral Large 3non-thinkingOpensame+297%$0.031Memory (4-bit): not verifiedMistral Small 4thinkingOpensame+92%$0.015Memory (4-bit): 1.9 GB → 72 GBMinistral 3 14B 2512non-thinkingOpensame+148%$0.019Memory (4-bit): 1.9 GB → 7 GBMinistral 3 8B 2512Opensame+41%$0.011Memory (4-bit): 1.9 GB → 4.5 GBMistral Smalln/a—$1 / $3Context: 131K → 32KMinistral 14B 3.0n/a—$0.20 / $0.20Context: 131K → 128KAll Mistral models →Run it yourself
estimateBenchmarks
in Agents & Tool UseFAQ
How much does Ministral 3 3B 2512 cost?›
$0.10 per 1M input tokens and $0.10 per 1M output tokens; cached input $0.01 per 1M. A typical benchmark task costs about $0.008 at default effort.
Is Ministral 3 3B 2512 free to use?›
There is no free API; the cheapest paid price is $0.10 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Ministral 3 3B 2512, and what is it best at?›
#262 of 334 in Chat & Knowledge, #75 of 132 in Coding, #116 of 133 in Agents & Tool Use, #220 of 245 in Writing & Language, #119 of 124 in Vision & Documents.
Is Ministral 3 3B 2512 better than Ring 2.6 1T?›
On AutomationBench-AA, Ministral 3 3B 2512 scores 0.7 vs 0.7 — about the same. It is 97% cheaper per task.
How can I use Ministral 3 3B 2512?›
Through the API of Mistral.
What is the context window of Ministral 3 3B 2512?›
131K tokens, with up to 105K output tokens.
When was Ministral 3 3B 2512 released? Is there a newer version?›
Released 2025-12-02. It is the current version in its line.
Is Ministral 3 3B 2512 open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Ministral 3 3B 2512 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Ministral 3 3B 2512 read images?›
It accepts text and images as input.
What is Ministral 3 3B 2512?›
Ministral 3 3B 2512 is an open-weights language model from Mistral, released Dec 2025.
What are the best alternatives to Ministral 3 3B 2512?›
Ling 3.0 Flash VL (cheaper), GPT-6 Luna (better), K2 Horizon 7B (open weights).