Mistral AIMagistral Small 1.2
Magistral Small 1.2 is a Mistral AI open-weights reasoning model, released 18 Sept 2025. Its best configuration ranks #659 of 785 on the BenchLeader Index at 40.3 ±5.1, in the lower half. The ± is the point: 279 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 instruction following (47) and lowest in knowledge (28). At $0.750 per million tokens blended it is mid-priced. Output speed of 78 tokens per second puts it slower than most, with a first answer in 27.1 s. Last measured 13 Oct 2026.
- Blended price
- $0.750/M
- $0.500 in · $1.50 out
- Output speed
- 78 tok/s
- measured by Artificial Analysis
- First answer
- 27 s
- Context
- 128k
- Full answer
- 33 s
- median, reasoning included
- Cost per run
- –
- Released
- 18 Sept 2025
- Overall index40
- Reasoning39
- Coding46
- Agents & tools34
- Maths46
- Knowledge28
- Instruction following47
- Multimodal38
- Long context34
- Composite39
Versions
Mistral AI has shipped 3 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 |
|---|---|---|---|
| Magistral Small 1.2this page | 18 Sept 2025 | 40.3 | #659 |
| Magistral Small | 10 Jun 2025 | 37.6 | #727 |
| Magistral Small 1.0 | – | – | – |
How its index has moved
27 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 |
|---|---|---|
| GPQA Diamond | 47.6%#224 | Epoch AI Benchmarking Hub |
| GPQA Diamond (AA)not in index | 66.3%#334 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 6.4%#389 | Artificial Analysis |
| GPQA Diamond (Vals)not in index | 58.3%#107 | Vals AI |
| CritPt | 0.3%#271 | Artificial Analysis |
| DTBenchnot in index | 61.3%#182 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | not stated | Source |
|---|---|---|
| SciCode | 35.2%#192 | SciCode |
| LiveCodeBench | 72.1%#81 | Vals AI |
| Terminal-Bench 2.1 (AA)not in index | 4.5%#208 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source |
|---|---|---|
| Terminal-Bench Hard | 4.5%#284 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 27.8%#272 | Artificial Analysis |
| GDPval-AA v2.1not in index | 0.0%#124 | Artificial Analysis |
| τ³-Banking (AA) | 4.5%#191 | Artificial Analysis |
| AutomationBenchnot in index | 0.2%#219 | Artificial Analysis |
Maths
| Benchmark | not stated | Source |
|---|---|---|
| OTIS Mock AIME | 28.1%#211 | Epoch AI Benchmarking Hub |
| AIME (Vals) | 80.7%#47 | Vals AI |
| MGSM | 86.3%#59 | Vals AI |
Knowledge
| Benchmark | not stated | Source |
|---|---|---|
| AA-Omniscience | -65.1#499 | Artificial Analysis |
| MMLU-Pro | 62.1%#127 | Vals AI |
| LegalBench | 40.0%#142 | Vals AI |
| CorpFin | 44.0%#112 | Vals AI |
| TaxEval | 60.3%#123 | Vals AI |
| MedQA | 82.4%#63 | Vals AI |
| AA-Omniscience: accuracynot in index | 13.4%#456 | Artificial Analysis |
| AA-Omniscience: non-hallucinationnot in index | 9.2%#432 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source |
|---|---|---|
| IFBench | 44.4%#213 | Artificial Analysis |
Multimodal
| Benchmark | not stated | Source |
|---|---|---|
| MMMU-Pro | 55.5%#226 | Artificial Analysis |
| MMMU-Pro (Vals)not in index | 65.2%#74 | Vals AI |
| MortgageTaxnot in index | 62.1%#55 | Vals AI |
Long context
| Benchmark | not stated | Source |
|---|---|---|
| AA-LCR | 19.3%#438 | Artificial Analysis |
Composite
| Benchmark | not stated | Source |
|---|---|---|
| AA Intelligence Index v4.3.2 | 8.6#433 | Artificial Analysis |
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 | $0.0007 | – | 30.9 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0069 | – | 34.7 s |
| Code edit | 6,000 / 1,500 | $0.0053 | – | 46.2 s |
| Agentic coding session | 60,000 / 4,000 | $0.036 | – | 1.3 min |
| Structured extraction | 2,000 / 200 | $0.0013 | – | 29.6 s |
More from Mistral AI
See also
Questions people ask
- What are Magistral Small 1.2's benchmark scores?
- As of 25 Oct 2026, Magistral Small 1.2 scores 40.3 ±5.1 on the BenchLeader Index, strongest in instruction following (47) and weakest in knowledge (28). That rests on 19 independent benchmark results, and every score on this page links to its publisher.
- How does Magistral Small 1.2 rank?
- Magistral Small 1.2 is #659 of 785 ranked configurations on 25 Oct 2026. 279 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 Magistral Small 1.2 cost per 1M tokens?
- As of 25 Oct 2026, Magistral Small 1.2 costs $0.500 per 1M input tokens and $1.50 per 1M output tokens, $0.750 blended (three input tokens per output token).
- When was Magistral Small 1.2 released?
- Mistral AI released Magistral Small 1.2 on 18 Sept 2025.
Data as of 25 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Magistral Small 1.2: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/magistral-small-1-2.