Model comparison
GPT-4.1 nano vs Mistral Medium
Mistral Medium is the stronger model overall, scoring 36.3 to 27.9 on the Noometry Index. GPT-4.1 nano costs 17× less per token, which makes it the better buy when Mistral Medium's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
Mistral AISummary
- They share 26 benchmarks with published results for both. GPT-4.1 nano scores higher in 0 categories and Mistral Medium in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Mistral Medium leads 60.0 to 40.5.
- The biggest single-benchmark swing is WeirdML: 19% for GPT-4.1 nano and 43.7% for Mistral Medium.
- GPT-4.1 nano is cheaper at $0.10 / $0.40 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium.
- GPT-4.1 nano accepts more context: 1.05M tokens versus 262K.
- Mistral Medium has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 nano | Mistral Medium | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 27.9 | 36.3 |
| Released | 2025-04-14 | 2023-12-11 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 33K | 262K |
| Input $ / M tokens | $0.10 | $1.50 |
| Output $ / M tokens | $0.40 | $7.50 |
| Results tracked | 38 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Medium leads
GPT-4.1 nano: 24.1 (#330), Mistral Medium: 34.2 (#243)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| SciCode | 25.9% | 40.2% |
| WeirdML | 19% | 43.7% |
| LMArena Coding | 1306 | 1434 |
| FrontierCode | — | 8% |
| Aider Polyglot | 8.9% | — |
| ALE-Bench | — | 763.98 |
Agentic & Tool Use Mistral Medium leads
GPT-4.1 nano: 26.5 (#104), Mistral Medium: 28.3 (#90)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| Berkeley Function Calling Leaderboard | 33% | 37.7% |
Reasoning Mistral Medium leads
GPT-4.1 nano: 8.5 (#349), Mistral Medium: 24.0 (#167)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| Kagi LLM Benchmark | 33.3% | 50% |
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1286 | 1426 |
| DTBench | 52.5% | 75.5% |
| LMCA | 5.5% | 26.1% |
| ARC-AGI-2 | 0% | — |
| ARC-AGI-1 | 0% | — |
| Surface Evolver Bench | — | 26.9% |
| Epoch Capabilities Index | 129.62 | — |
Math Mistral Medium leads
GPT-4.1 nano: 26.9 (#252), Mistral Medium: 28.1 (#245)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 28.9% | 32.2% |
| LMArena Math | 1274 | 1408 |
| MATH Level 5 | 70% | 81.6% |
| FrontierMath (Feb 2025 set) | 1% | 0.3% |
| ProofBench | — | 9% |
| Omni-MATH | 36.7% | — |
Knowledge Mistral Medium leads
GPT-4.1 nano: 21.8 (#273), Mistral Medium: 25.0 (#265)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| GPQA Diamond | 48.9% | 59.5% |
| LMArena Expert | 1272 | 1408 |
| Humanity's Last Exam | — | 4.5% |
| SimpleQA Verified | 6% | — |
| MMLU-Pro | 55% | — |
| Vectara Hallucination Rate | — | 22.7% |
| GPQA (HELM) | 50.7% | — |
Multimodal Mistral Medium leads
GPT-4.1 nano: 29.2 (#113), Mistral Medium: 35.3 (#88)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| LMArena Vision | 1063 | 1172 |
Multilingual Mistral Medium leads
GPT-4.1 nano: 41.6 (#205), Mistral Medium: 52.1 (#91)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| LMArena Non-English | 1260 | 1408 |
| LMArena Chinese | 1270 | 1447 |
| LMArena German | 1288 | 1432 |
| LMArena Japanese | 1198 | 1378 |
| LMArena Russian | 1261 | 1411 |
| LMArena French | — | 1459 |
| LMArena Korean | — | 1380 |
| LMArena Spanish | — | 1433 |
Instruction Following Mistral Medium leads
GPT-4.1 nano: 67.8 (#193), Mistral Medium: 73.7 (#116)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| LMArena Instruction Following | 1267 | 1398 |
| IFEval | 84.3% | — |
Long Context Mistral Medium leads
GPT-4.1 nano: 23.7 (#296), Mistral Medium: 42.9 (#114)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| LMArena Longer Query | 1283 | 1406 |
| Fiction.LiveBench | 25% | — |
Writing & Preference Mistral Medium leads
GPT-4.1 nano: 40.5 (#243), Mistral Medium: 60.0 (#103)
| Benchmark | GPT-4.1 nano | Mistral Medium |
|---|---|---|
| LMArena Text | 1285 | 1424 |
| LMArena Creative Writing | 1260 | 1391 |
| LMArena Multi-Turn | 1277 | 1418 |
| Short-Story Creative Writing | — | 77.3% |
| EQ-Bench Creative Writing | 946 | — |
| WildBench | 81.2% | — |
Frequently asked questions
Is GPT-4.1 nano better than Mistral Medium?
Mistral Medium is the stronger model overall, scoring 36.3 to 27.9 on the Noometry Index. GPT-4.1 nano costs 17× less per token, which makes it the better buy when Mistral Medium's lead doesn't matter for your workload.
Which is cheaper, GPT-4.1 nano or Mistral Medium?
GPT-4.1 nano is cheaper. It lists at $0.10 per million input tokens and $0.40 per million output tokens; Mistral Medium lists at $1.50 and $7.50.
Is GPT-4.1 nano or Mistral Medium better for coding?
Mistral Medium scores higher on coding benchmarks: 34.2 versus 24.1 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 nano does, with 1.05M tokens against 262K.
How many benchmarks do GPT-4.1 nano and Mistral Medium share?
26 benchmarks have published results for both models. GPT-4.1 nano has 38 scored results on Noometry and Mistral Medium has 36.