Model comparison
DeepSeek-V3.1 vs Magistral Medium
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 35.2 on the Noometry Index.
Last verified . 18 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 18 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 7 categories and Magistral Medium in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek-V3.1 leads 27.9 to 8.6.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 53.2% for DeepSeek-V3.1 and 16.2% for Magistral Medium.
- DeepSeek-V3.1 is cheaper at $0.25 / $0.95 per million input/output tokens, against $2 / $5 for Magistral Medium.
- Magistral Medium accepts more context: 262K tokens versus 164K.
Side by side
| DeepSeek-V3.1 | Magistral Medium | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 42.8 | 35.2 |
| Released | 2025-08-21 | 2025-03-17 |
| Weights | Open | Open |
| Context window | 164K | 262K |
| Max output | 8K | 16K |
| Input $ / M tokens | $0.25 | $2 |
| Output $ / M tokens | $0.95 | $5 |
| Results tracked | 27 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1 leads
DeepSeek-V3.1: 40.3 (#144), Magistral Medium: 39.1 (#161)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Coding | 1417 | 1319 |
| SciCode | — | 39.2% |
| WeirdML | 38.4% | — |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), Magistral Medium: 8.6 (#348)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | 16.2% |
| LMArena Hard Prompts | 1417 | 1267 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | 40% | — |
| ARC-AGI-1 | — | 6.1% |
| CritPt | — | 0.3% |
| DTBench | 82.7% | — |
| LMCA | 24.3% | — |
| Epoch Capabilities Index | 139.92 | — |
| ForecastBench | 58 | — |
Math DeepSeek-V3.1 leads
DeepSeek-V3.1: 38.9 (#122), Magistral Medium: 35.1 (#189)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Math | 1420 | 1250 |
Knowledge DeepSeek-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), Magistral Medium: 33.5 (#202)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Expert | 1405 | 1223 |
| Vectara Hallucination Rate | 5.5% | — |
Multilingual DeepSeek-V3.1 leads
DeepSeek-V3.1: 51.6 (#106), Magistral Medium: 39.6 (#224)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Non-English | 1400 | 1232 |
| LMArena Chinese | 1469 | 1227 |
| LMArena French | 1447 | 1267 |
| LMArena German | 1411 | 1248 |
| LMArena Japanese | 1378 | 1175 |
| LMArena Korean | 1337 | 1125 |
| LMArena Russian | 1405 | 1224 |
| LMArena Spanish | 1431 | 1271 |
Instruction Following DeepSeek-V3.1 leads
DeepSeek-V3.1: 73.9 (#110), Magistral Medium: 66.0 (#211)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Instruction Following | 1400 | 1254 |
Long Context Magistral Medium leads
DeepSeek-V3.1: 36.3 (#232), Magistral Medium: 39.3 (#183)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Longer Query | 1422 | 1295 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference DeepSeek-V3.1 leads
DeepSeek-V3.1: 60.3 (#98), Magistral Medium: 46.3 (#219)
| Benchmark | DeepSeek-V3.1 | Magistral Medium |
|---|---|---|
| LMArena Text | 1420 | 1255 |
| LMArena Creative Writing | 1401 | 1245 |
| LMArena Multi-Turn | 1408 | 1275 |
| EQ-Bench Creative Writing | 1436 | — |
Frequently asked questions
Is DeepSeek-V3.1 better than Magistral Medium?
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 35.2 on the Noometry Index.
Which is cheaper, DeepSeek-V3.1 or Magistral Medium?
DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Magistral Medium lists at $2 and $5.
Is DeepSeek-V3.1 or Magistral Medium better for coding?
DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 39.1 in the Noometry coding category.
Which has the bigger context window?
Magistral Medium does, with 262K tokens against 164K.
How many benchmarks do DeepSeek-V3.1 and Magistral Medium share?
18 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Magistral Medium has 22.