Mistral AI model product
Magistral Medium
Magistral Medium is a reasoning model trained with reinforcement learning on top of Mistral Medium 3 for complex math and code tasks.
Updated Oct 23, 2026. Default version: Magistral Medium
LLMBoard Score16.6Magistral Medium
Coverage60%6 benchmark families
Context windowN/ATokens
Official input priceN/AOfficial price unavailable
Magistral Medium Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Magistral Medium Benchmark Results
Benchmark scores for Magistral Medium.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkAider-Polyglot | Score47.10% | Rank18 | Participants22 | Percentile19.05% | EvidenceC | Evaluated |
| BenchmarkAIME 2024 | Score73.60% | Rank37 | Participants54 | Percentile32.08% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench | Score50.30% | Rank47 | Participants76 | Percentile38.67% | EvidenceC | Evaluated |
| BenchmarkAA AIME 2025 | Score40.33% | Rank63 | Participants96 | Percentile34.74% | EvidenceB | EvaluatedN/A |
| BenchmarkHumanity's Last Exam | Score9.00% | Rank99 | Participants109 | Percentile9.26% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score64.90% | Rank100 | Participants122 | Percentile18.18% | EvidenceC | Evaluated |
| BenchmarkAA CritPt | Score0.29% | Rank118 | Participants200 | Percentile41.21% | EvidenceB | EvaluatedN/A |
| BenchmarkAA Omniscience Accuracy | Score20.33% | Rank121 | Participants201 | Percentile40.00% | EvidenceB | EvaluatedN/A |
| BenchmarkAA HLE Text No Tools | Score9.82% | Rank128 | Participants200 | Percentile36.18% | EvidenceB | EvaluatedN/A |
| BenchmarkAA GPQA Diamond | Score67.88% | Rank138 | Participants199 | Percentile30.81% | EvidenceB | EvaluatedN/A |
| BenchmarkGPQA | Score70.80% | Rank145 | Participants256 | Percentile43.53% | EvidenceC | Evaluated |
| BenchmarkAA IFBench | Score25.10% | Rank161 | Participants168 | Percentile4.19% | EvidenceB | EvaluatedN/A |
| BenchmarkAA LCR v1.1 | Score0.00% | Rank185 | Participants193 | Percentile4.17% | EvidenceB | EvaluatedN/A |
| BenchmarkLM Arena Text Style Control | Score1,304.92 rating | Rank187 | Participants218 | Percentile14.29% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text | Score1,254.93 rating | Rank194 | Participants218 | Percentile11.06% | EvidenceA | Evaluated |
Magistral Medium Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext style control | Categorypolish | Rank147 | Rating / score1,312.33 | Votes1,619 | ObservationsN/A | Result date |
| Arenatext style control | Categorycoding | Rank155 | Rating / score1,387.30 | Votes2,245 | ObservationsN/A | Result date |
| Arenatext style control | Categorygerman | Rank155 | Rating / score1,306.92 | Votes360 | ObservationsN/A | Result date |
| Arenatext | Categorypolish | Rank156 | Rating / score1,248.08 | Votes1,619 | ObservationsN/A | Result date |
| Arenatext style control | Categoryspanish | Rank159 | Rating / score1,319.97 | Votes269 | ObservationsN/A | Result date |
| Arenatext | Categoryjapanese | Rank160 | Rating / score1,175.51 | Votes349 | ObservationsN/A | Result date |
| Arenatext style control | Categoryjapanese | Rank161 | Rating / score1,224.37 | Votes349 | ObservationsN/A | Result date |
| Arenatext | Categoryspanish | Rank163 | Rating / score1,270.50 | Votes269 | ObservationsN/A | Result date |
| Arenatext style control | Categoryfrench | Rank163 | Rating / score1,325.19 | Votes184 | ObservationsN/A | Result date |
| Arenatext | Categoryfrench | Rank165 | Rating / score1,267.16 | Votes184 | ObservationsN/A | Result date |
| Arenatext | Categorycoding | Rank166 | Rating / score1,318.88 | Votes2,245 | ObservationsN/A | Result date |
| Arenatext style control | Categorylonger query | Rank166 | Rating / score1,336.49 | Votes2,402 | ObservationsN/A | Result date |
| Arenatext | Categorygerman | Rank168 | Rating / score1,248.27 | Votes360 | ObservationsN/A | Result date |
| Arenatext | Categorylonger query | Rank171 | Rating / score1,294.28 | Votes2,402 | ObservationsN/A | Result date |
| Arenatext style control | Categoryhard prompts english | Rank171 | Rating / score1,351.77 | Votes2,395 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry entertainment and sports and media | Rank171 | Rating / score1,287.64 | Votes1,972 | ObservationsN/A | Result date |
| Arenatext style control | Categoryhard prompts | Rank172 | Rating / score1,332.80 | Votes5,640 | ObservationsN/A | Result date |
| Arenatext style control | Categorykorean | Rank174 | Rating / score1,190.30 | Votes194 | ObservationsN/A | Result date |
| Arenatext | Categorykorean | Rank175 | Rating / score1,124.82 | Votes194 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry software and it services | Rank175 | Rating / score1,351.59 | Votes4,007 | ObservationsN/A | Result date |
| Arenatext style control | Categoryinstruction following | Rank177 | Rating / score1,300.78 | Votes3,115 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts english | Rank179 | Rating / score1,284.36 | Votes2,395 | ObservationsN/A | Result date |
| Arenatext style control | Categorymulti turn | Rank179 | Rating / score1,313.44 | Votes1,599 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry medicine and healthcare | Rank180 | Rating / score1,334.39 | Votes775 | ObservationsN/A | Result date |
| Arenatext style control | Categorymath | Rank180 | Rating / score1,286.85 | Votes551 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank182 | Rating / score1,266.89 | Votes5,640 | ObservationsN/A | Result date |
| Arenatext style control | Categorycreative writing | Rank182 | Rating / score1,280.15 | Votes1,350 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry mathematical | Rank182 | Rating / score1,293.74 | Votes571 | ObservationsN/A | Result date |
| Arenatext | Categorycreative writing | Rank183 | Rating / score1,244.88 | Votes1,350 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry entertainment and sports and media | Rank183 | Rating / score1,241.70 | Votes1,972 | ObservationsN/A | Result date |
Magistral Medium Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $2 input, $5 output per 1M via Pioneer
- Tracked offerings
- 1
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderPioneer | Provider model IDmagistral-medium | Regionglobal | Input / 1M$2 | Output / 1M$5 | Context128K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Magistral Medium Runtime Performance
Provider-specific output speed and catalog latency for Magistral Medium. Runtime does not affect the capability score.
No runtime data
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Magistral Medium Specifications
Technical details for the model's default version.
- Version
- Magistral Medium
- Released
- Jun 10, 2025
- Knowledge cutoff
- Jun 1, 2025
- Parameters
- 24B
- Context window
- N/A
- Max output
- N/A
- Inputs
- image, text
- Outputs
- text
- Open weights
- No
- License
- Apache 2.0
Magistral Medium Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionMagistral Medium | Released | LLMBoard16.64 | Parameters24B | ContextN/A | Max outputN/A | Open weightsNo | LicenseApache 2.0 |
Models similar to Magistral Medium
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
MA
Magistral Small
Mistral AI
MA
Min istral 3
Mistral AI
MA
Mistral Large 3
Mistral AI
MA
Devstral Small 1.1
Mistral AI
MA
Pixtral Large
Mistral AI
MA
Mistral Small 3.2 24B
Mistral AI
What is Magistral Medium?
Key information about Magistral Medium and its available data.
Magistral Medium is a reasoning model from Mistral AI, trained solely with reinforcement learning on top of Mistral Medium 3 without distillation from existing reasoning models. Its training uses an RLVR framework with modifications to GRPO to improve reasoning ability and multilingual consistency.
Data as of 2026-10-22.
FAQ
Common questions about Magistral Medium.
When was Magistral Medium released?
Magistral Medium's default version was released on Jun 10, 2025.
How much does Magistral Medium cost?
No official standard PAYG price is currently available for Magistral Medium. The lowest tracked third-party offer starts at $2 input and $5 output via Pioneer.
Who created Magistral Medium?
Magistral Medium was created by Mistral AI.
What is the context window for Magistral Medium?
A context window is not available for the default version.
Is Magistral Medium open weight?
No. The default version is not marked as having publicly available weights.