Mistral AIMistral Small 3.2
Mistral Small 3.2 is a Mistral AI open-weights model, released 20 Jun 2025. Its best configuration ranks #622 of 785 on the BenchLeader Index at 41.6 ±3.8, in the lower half. The ± is the point: 288 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (52) and lowest in long context (34). At $0.110 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 42 tokens per second puts it in the slowest quarter, with a first answer in 1.1 s. Last measured 22 Oct 2026.
- Blended price
- $0.110/M
- $0.080 in · $0.200 out
- Output speed
- 42 tok/s
- measured by Artificial Analysis
- First answer
- 1.08 s
- first token 0.51 s
- Context
- 128k
- Full answer
- 13 s
- median, reasoning included
- Cost per run
- $0.109
- one full Intelligence Index run
- Released
- 20 Jun 2025
- Overall index42
- Reasoning41
- Coding37
- Agents & tools41
- Maths46
- Knowledge39
- Instruction following37
- Human preference52
- Multimodal37
- Long context34
- Composite37
Versions
Mistral AI has shipped 6 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Mistral Small 4thinking | 16 Mar 2026 | 45.6 | #511 |
| Mistral Small 3.2this page | 20 Jun 2025 | 41.6 | #622 |
| Mistral Small 3.1 | 17 Mar 2025 | 36.3 | #749 |
| Mistral Small 3 | 25 Jan 2025 | 39.3 | #688 |
| Mistral Small v24.09 | 17 Sept 2024 | – | – |
| Mistral Small 2402 | 26 Feb 2024 | 32.0 | #781 |
How its index has moved
9 Sept 2026 to 21 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| GPQA Diamond | 49.0%#218 | Epoch AI Benchmarking Hub | |
| Arena Hard Prompts | 1373#195 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 50.5%#428 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.3%#510 | Artificial Analysis | |
| Kagi LLM Benchmark | 40.4%#105 | Kagi LLM Benchmark | |
| CritPt | 0.0%#315 | Artificial Analysis | |
| DTBenchnot in index | 59.9%#187 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| SciCode | 26.4%#204 | SciCode | |
| Arena Coding | 1411#186 | Arena (LMArena) | |
| SciCode (AA)not in index | 28.6%#213 | Artificial Analysis | |
| Terminal-Bench 4.0 (AA)not in index | 0.0%#152 | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 5.6%#203 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench Hard | 6.8%#249 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 29.5%#261 | Artificial Analysis | |
| BFCL Overall | 37.1%#39 | Berkeley Function Calling Leaderboard | |
| τ³-Banking (AA) | 6.2%#169 | Artificial Analysis | |
| AutomationBenchnot in index | 0.3%#213 | Artificial Analysis | |
| GDP.pdf (AA)not in index | 0.0%#220 | Artificial Analysis |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| OTIS Mock AIME | 30.3%#207 | Epoch AI Benchmarking Hub | |
| Arena Maths | 1337#207 | Arena (LMArena) |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -50.6#390 | Artificial Analysis | |
| AA-Omniscience: accuracynot in index | 14.8%#429 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 23.2%#257 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 33.5%#327 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1336#202 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1356#194 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1324#191 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1364#184 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1358#199 | Arena (LMArena) |
Multimodal
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Vision | 1142#118 | Arena (LMArena) | |
| MMMU-Pro | 48.0%#249 | Artificial Analysis |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR | 20.3%#432 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index v4.3.2 | 7.0#498 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Mistral | 141 tok/s | 1.74 s | $0.100 | $0.300 | 33k | – |
| DeepInfra | 46 tok/s | 0.52 s | $0.075 | $0.200 | 128k | fp8 |
| Parasail | 45 tok/s | 0.69 s | $0.090 | $0.300 | 131k | bf16 |
| Venice | 24 tok/s | 0.66 s | $0.094 | $0.250 | 256k | fp8 |
Price history
The price above, per 1M tokens, every day it changed.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0001 | – | 8.2 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0011 | – | 15.4 s |
| Code edit | 6,000 / 1,500 | $0.0008 | – | 36.9 s |
| Agentic coding session | 60,000 / 4,000 | $0.0056 | – | 1.6 min |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 5.9 s |
More from Mistral AI
See also
Questions people ask
- What are Mistral Small 3.2's benchmark scores?
- As of 24 Oct 2026, Mistral Small 3.2 scores 41.6 ±3.8 on the BenchLeader Index, strongest in human preference (52) and weakest in long context (34). That rests on 18 independent benchmark results, and every score on this page links to its publisher.
- How does Mistral Small 3.2 rank?
- Mistral Small 3.2 is #622 of 785 ranked configurations on 24 Oct 2026. 288 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Mistral Small 3.2 cost per 1M tokens?
- As of 24 Oct 2026, Mistral Small 3.2 costs $0.080 per 1M input tokens and $0.200 per 1M output tokens, $0.110 blended (three input tokens per output token). The list price last changed on 6 Oct 2026.
- When was Mistral Small 3.2 released?
- Mistral AI released Mistral Small 3.2 on 20 Jun 2025.
Data as of 24 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Mistral Small 3.2: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/mistral-small-3-2.