Mistral Small
Budgetby Mistral
The mistral-small-latest alias now points to Mistral Small 4 (mistral-small-2603), released March 16, 2026. Mistral describes it as a hybrid model that unifies instruct, reasoning, and coding in a single set of weights, with 119 billion total parameters and 6.5 billion active, Apache 2.0 licensing, a 256K context window, and pricing of $0.15 per million input tokens and $0.60 output. Mistral Small 3.2 was retired on July 31, 2026, so anything still pinned to an older Small snapshot needs a migration test.
Input Price
$0.15
per 1M tokens
Output Price
$0.60
per 1M tokens
Context Window
262K
tokens
Released
2026-03
Open source
Capabilities
Key Strengths
- ✓119B total, 6.5B active
- ✓Apache 2.0 open weights
- ✓Instruct, reasoning, and coding in one model
- ✓256K context window
- ✓Tool use support
Best For
- ▸High-volume chat
- ▸Lightweight code tasks
- ▸Quick summarization
- ▸Classification
Benchmark Scores
| Benchmark | Score | Description |
|---|---|---|
| MMLU-Pro | 78.4 | General knowledge and reasoning across 57 subjects |
| HumanEval | 82.5 | Python code generation and problem solving |
| GPQA Diamond | 44.6 | Graduate-level science questions verified by domain experts |
| MATH | 68.9 | Competition-level mathematics problems |
Scores sourced from public benchmark datasets. See full benchmark leaderboard for all models.
Pricing Details
Input tokens
$0.15
per 1M tokens
Output tokens
$0.60
per 1M tokens
Estimated cost per 1K requests
$0.45
~1K input + ~500 output tokens avg
Prices are subject to change. Check the official documentation for current pricing. See the cost calculator for detailed estimates.
Open Source Model
Mistral Small is free to download and self-host under the Apache-2.0. Hosted API pricing varies by provider (e.g., Together, Fireworks, Groq). See our open source LLM guide for deployment options.