Ministral 3 3B
MistralMistralOpen WeightApache 2.0 · Commercial OK
Description
The smallest model in the Ministral 3 family, Ministral 3 3B is a powerful, efficient tiny language model with vision capabilities. This model is the base pre-trained version, not fine-tuned for instruction or reasoning tasks, making it ideal for custom post-training processes. The Ministral 3 family is designed for edge deployment, capable of running on a wide range of hardware. Ministral 3 3B can even be deployed locally, fitting in 16GB of VRAM in BF16, and less than 8GB of RAM/VRAM when quantized.
Release Date
2025-12-02
Parameters
3.0B
Context Length
131K
Modalities
image, text
Capability Radar
23
general
13
coding
24
reasoning
22
scienceest.
0
agents
10
multimodal
Science uses a reasoning proxy when dedicated science benchmarks are unavailable.
Rankings
| Domain | #Rank | Score | Source |
|---|---|---|---|
| Code Ranking | 407 | 12.0 | AA |
| General Ranking | 397 | 25.0 | AA |
| Math Reasoning | 296 | 22.0 | AA |
| Science | 413 | 21.0 | AA |
Benchmark Scores (LLM Stats)
Finance
MMLU
70.7%SR
General
MMLU-Redux
73.5%SR
Multilingual MMLU
65.2%SR
TriviaQA
59.2%SR
AGIEval
51.1%SR
Math
MATH (CoT)
60.1%SR
AA Evaluation Indices
Math Index22.0
Intelligence Index11.2
Coding Index4.8
Mmlu Pro0.5
Gpqa0.4
Ifbench0.3
Tau20.2
Livecodebench0.2
Aime 250.2
Scicode0.1
Lcr0.1
Hle0.1
Terminalbench Hard0.0
LLM Stats Category Scores
Finance70
Healthcare70
Language70
General60
Legal60
Math60
Reasoning60
Pricing
Input Price$0.1 / 1M tokens
Output Price$0.1 / 1M tokens
Blended Price (3:1)$0.1 / 1M tokens
Speed
Tokens/sec290.6 tokens/s
Time to First Token0.31s
Time to Answer0.31s
Available Providers
(LS internal units)No provider data available