Model comparison
Mistral Medium vs Mistral Small
Mistral Medium is the stronger model overall, scoring 36.3 to 33.4 on the Noometry Index. Mistral Small costs 11× less per token, which makes it the better buy when Mistral Medium's lead doesn't matter for your workload.
Last verified . 29 shared benchmarks.
Mistral AISummary
- They share 29 benchmarks with published results for both. Mistral Medium scores higher in 9 categories and Mistral Small in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Mistral Medium leads 28.1 to 16.4.
- The biggest single-benchmark swing is MATH Level 5: 81.6% for Mistral Medium and 46.8% for Mistral Small.
- Mistral Small is cheaper at $0.15 / $0.60 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium.
Side by side
| Mistral Medium | Mistral Small | |
|---|---|---|
| Provider | Mistral AI | Mistral AI |
| Noometry Index | 36.3 | 33.4 |
| Released | 2023-12-11 | 2024-02-26 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 262K | 256K |
| Input $ / M tokens | $1.50 | $0.15 |
| Output $ / M tokens | $7.50 | $0.60 |
| Results tracked | 36 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Mistral Medium: 34.2 (#243), Mistral Small: 34.0 (#247)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| SciCode | 40.2% | 26.5% |
| LMArena Coding | 1434 | 1362 |
| ALE-Bench | 763.98 | 497.62 |
| FrontierCode | 8% | — |
| WeirdML | 43.7% | — |
| BigCodeBench Instruct | — | 36.1% |
| LiveBench Coding | — | 36.2% |
| BigCodeBench Complete | — | 46.6% |
Agentic & Tool Use Too close to call
Mistral Medium: 28.3 (#90), Mistral Small: 28.1 (#93)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | 37.7% | 37.1% |
Reasoning Mistral Medium leads
Mistral Medium: 24.0 (#167), Mistral Small: 19.8 (#250)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| Kagi LLM Benchmark | 50% | 37.8% |
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1426 | 1335 |
| DTBench | 75.5% | 70.9% |
| LMCA | 26.1% | 20.6% |
| LiveBench Reasoning | — | 44.8% |
| LiveBench Data Analysis | — | 53.7% |
| Surface Evolver Bench | 26.9% | — |
| LiveBench | — | 44% |
Math Mistral Medium leads
Mistral Medium: 28.1 (#245), Mistral Small: 16.4 (#293)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 32.2% | 5.8% |
| LMArena Math | 1408 | 1341 |
| MATH Level 5 | 81.6% | 46.8% |
| ProofBench | 9% | — |
| LiveBench Math | — | 39.9% |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Mistral Small leads
Mistral Medium: 25.0 (#265), Mistral Small: 31.0 (#222)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| GPQA Diamond | 59.5% | 47.5% |
| Vectara Hallucination Rate | 22.7% | 5.1% |
| LMArena Expert | 1408 | 1291 |
| Humanity's Last Exam | 4.5% | — |
| MMLU | — | 68.7% |
Multimodal Mistral Medium leads
Mistral Medium: 35.3 (#88), Mistral Small: 33.5 (#96)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| LMArena Vision | 1172 | 1142 |
Multilingual Mistral Medium leads
Mistral Medium: 52.1 (#91), Mistral Small: 45.5 (#169)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| LMArena Non-English | 1408 | 1315 |
| LMArena Chinese | 1447 | 1340 |
| LMArena French | 1459 | 1337 |
| LMArena German | 1432 | 1340 |
| LMArena Japanese | 1378 | 1275 |
| LMArena Korean | 1380 | 1259 |
| LMArena Russian | 1411 | 1324 |
| LMArena Spanish | 1433 | 1346 |
Instruction Following Mistral Medium leads
Mistral Medium: 73.7 (#116), Mistral Small: 66.4 (#209)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| LMArena Instruction Following | 1398 | 1310 |
| LiveBench Instruction Following | — | 63.7% |
Long Context Mistral Medium leads
Mistral Medium: 42.9 (#114), Mistral Small: 40.4 (#156)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| LMArena Longer Query | 1406 | 1327 |
Writing & Preference Mistral Medium leads
Mistral Medium: 60.0 (#103), Mistral Small: 52.5 (#171)
| Benchmark | Mistral Medium | Mistral Small |
|---|---|---|
| LMArena Text | 1424 | 1338 |
| LMArena Creative Writing | 1391 | 1305 |
| LMArena Multi-Turn | 1418 | 1344 |
| Short-Story Creative Writing | 77.3% | — |
| LiveBench Language | — | 30.5% |
Frequently asked questions
Is Mistral Medium better than Mistral Small?
Mistral Medium is the stronger model overall, scoring 36.3 to 33.4 on the Noometry Index. Mistral Small costs 11× less per token, which makes it the better buy when Mistral Medium's lead doesn't matter for your workload.
Which is cheaper, Mistral Medium or Mistral Small?
Mistral Small is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Mistral Medium lists at $1.50 and $7.50.
Is Mistral Medium or Mistral Small better for coding?
They score almost the same on coding (34.2 vs 34.0); test both on your own repository before choosing.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Mistral Medium and Mistral Small share?
29 benchmarks have published results for both models. Mistral Medium has 36 scored results on Noometry and Mistral Small has 39.