Ctrl K
Back to models

Mistral

Ministral 3 8B

  • Released
  • Open weights
  • Apache 2.0
Compare
Intelligence Index

Standard setting

5.5#81 of 83measured models
Measured · Artificial Analysis v4.3.2
API price from
$0.15 in / $0.15 out / 1M tokOpenRouter · Mistral
Context window
256K tok

Performance profile

ReasoningRank #96 of 107
47.1%
GPQA Diamond
CodingRank #91 of 94
20.7%
SciCode
Cost efficiencyRank #8 of 67
494.1pts / $
Intelligence ÷ eval cost

Capabilities

Context window

Modalities

Measured setting

Integration

Reasoning trade-offs

Source (opens in a new tab)
Intelligence Index v4.3.2
Ministral 3 8B
Cost per task (USD) · log scale
Ministral 3 8BStandard
Intelligence Index
5.5
Cost / task
$0.0111
Evaluation time
1.3m

Pricing & access

API route
OpenRouter · MistralAPI
Tier & contextStandard

USD per 1M tokens

Input
$0.15
Output
$0.15
Cached input
$0.015

Run locally

GGUF
Context
Estimated memory

~7.9 GiB

RAM + VRAM

Estimate assumptions

One text chat at the selected context, with a 16-bit cache and app allowance. Operating system and image processing are extra; optimized caches can use less.

16 GiB RAMwithout GPU offload

Memory budgetOne text chat
Weights
4.8 GiB
Context cache
1.1 GiB16-bit cache
App allowance
2 GiB
Disk space
~12 GBIncludes image file
Available software
Ollama (opens in a new tab)LM Studio (opens in a new tab)Jan (opens in a new tab)KoboldCpp (opens in a new tab)llama.cpp (opens in a new tab)
Get GGUF (opens in a new tab)
Set up in Ollama · Q4_K_M · 8K context

Save the following as Modelfile. The context matches the estimate above.

FROM hf.co/unsloth/Ministral-3-8B-Instruct-2512-GGUF:Q4_K_M
PARAMETER num_ctx 8192

Then run these commands in your terminal:

ollama create llmcurrent-local -f Modelfile
ollama run llmcurrent-local

Ollama import instructions (opens in a new tab) · For LM Studio or Jan, open the app’s model search, enter unsloth/Ministral-3-8B-Instruct-2512-GGUF, choose Q4_K_M, and set context to 8,192 tokens in the app’s model settings. Software links above open each app’s setup guide.

Images need extra memoryvLLM setup (opens in a new tab)

Benchmarks

Full leaderboardSource (opens in a new tab)

Humanity’s Last Exam

Frontier reasoning

Ministral 3 8B · Standard#120 of 126

  1. 1Claude Opus 5.5Max61.4%
  2. 2Claude Fable 5.1Max59.1%
  3. 3Gemini 4 ArgonHigh57.1%
  5. 117G9v3-3BReasoning4.5%
  6. 118Phi-4 Mini InstructStandard4.5%
  7. 119Reka Flash 3Reasoning4.4%
  8. 120Ministral 3 8BStandard4.3%
  9. 121Nova PremierStandard4.2%
  10. 122Command AStandard4.0%
  11. 123Granite 4.0 H SmallStandard3.8%

Peers use their best measured result on this benchmark.

Alternatives

Each model at its highest-scoring measured setting · Intelligence Index v4.3.2 and its evaluation cost per task.