Mistral Small 3
by Mistral · mistralai/mistral-small-24b-instruct-2501 · #14 cheapest of 325 paid models
Prices updated Jul 22, 2026, 2:18 AM UTC · refreshed hourly
Mistral Small 3 is a 24B-parameter language model optimized for low-latency performance across common AI tasks. Released under the Apache 2.0 license, it features both pre-trained and instruction-tuned versions designed...
Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol
Pricing
Input
$0.05
per 1M tokens
Output
$0.08
per 1M tokens
Blended (3:1)
$0.057
3 input : 1 output
Cache read
—
per 1M cached tokens
Cache write
—
per 1M tokens
Cache & batch economics
Effective prices for Mistral Small 3 (no cache-read rate published — hits billed as normal input).
| Cache hit rate | Effective blended $/1M | Example request* | vs no cache |
|---|---|---|---|
| 0% | $0.057 | $0.00518 | — |
| 50% | $0.057 | $0.00518 | −0% |
| 90% | $0.057 | $0.00518 | −0% |
*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates (available for OpenAI / Anthropic / Google — not flagged for this creator). Confirm on the vendor’s pricing page.
What a request costs
| Workload | Input tokens | Output tokens | Cost / request | Cost / 1K requests |
|---|---|---|---|---|
| Short chat message | 500 | 300 | $4.90e-5 | $0.049 |
| Document summary | 8,000 | 1,000 | $0.00048 | $0.48 |
| Codebase question (RAG) | 30,000 | 2,000 | $0.00166 | $1.66 |
| Long-context analysis | 150,000 | 5,000 | $0.0079 | $7.90 |
Performance
Intelligence Index
6.9
AA composite quality
Coding Index
—
Math Index
4.3
Agentic Index
—
Output speed
158 tok/s
median
Time to first token
0.53s
median
Time to first answer
0.53s
after reasoning tokens
Artificial Analysis benchmarks
| Benchmark | Score |
|---|---|
| MMLU-Pro | 65.2% |
| GPQA Diamond | 46.2% |
| Humanity's Last Exam | 4.1% |
| LiveCodeBench | 25.2% |
| SciCode | 23.6% |
| MATH-500 | 71.5% |
| AIME | 8.0% |
| AIME 2025 | 4.3% |
| IFBench | 26.4% |
| AA-LCR (Long Context Reasoning) | 0.0% |
| τ²-Bench (Telecom) | 19.6% |
Specs
Context window
33K
tokens
Max output
16K
tokens
Input modalities
text
Tokenizer
Mistral