Ctrl K
Back to models

Mistral

Mistral Small 4

  • Released
  • Open weights
  • Apache 2.0
Compare
Intelligence Index

Reasoning setting

11.3#67 of 83measured models
Measured · Artificial Analysis v4.3.2
API price from
$0.15 in / $0.60 out / 1M tokOpenRouter · Mistral
Context window
256K tok

Performance profile

ReasoningRank #59 of 107
76.9%
GPQA Diamond
CodingRank #69 of 94
38.8%
SciCode
Cost efficiencyRank #3 of 67
748.8pts / $
Intelligence ÷ eval cost

Capabilities

Context window

Modalities

Measured setting

Integration

Reasoning trade-offs

Source (opens in a new tab)
Intelligence Index v4.3.2
Mistral Small 4
Cost per task (USD) · log scale
Mistral Small 4Reasoning
Intelligence Index
11.3
Cost / task
$0.015
Evaluation time
1.3m

Pricing & access

API route
OpenRouter · MistralAPI
Tier & contextStandard

USD per 1M tokens

Input
$0.15
Output
$0.60
Cached input
$0.015

Run locally

GGUF
Context
Estimated memory

~78.4 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.

96 GiB RAMwithout GPU offload

Memory budgetOne text chat
Weights
67.2 GiB
Context cache
4.5 GiB16-bit cache
App allowance
6.7 GiB
Download
72.2 GB2 files · keep together+0.9 GB for images (opens in a new tab)
Disk space
~86 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/lmstudio-community/Mistral-Small-4-119B-2603-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 lmstudio-community/Mistral-Small-4-119B-2603-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

Mistral Small 4 · Reasoning#93 of 126

  1. 1Claude Opus 5.5Max61.4%
  2. 2Claude Fable 5.1Max59.1%
  3. 3Gemini 4 ArgonHigh57.1%
  5. 90Nemotron 3.5 LightningReasoning10.6%
  6. 91Qwen3 Coder NextStandard10.1%
  7. 92Qwen3.5 4BReasoning9.9%
  8. 93Mistral Small 4Reasoning9.9%
  9. 94Granite 4.2 8BReasoning9.7%
  10. 95Nova 2.0 Pro PreviewMedium9.4%
  11. 96Ling 3.0 TinyReasoning9.3%

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.