Model comparison
GPT-4o vs Mistral Large
Mistral Large is the stronger model overall, scoring 31.9 to 28.6 on the Noometry Index.
Last verified . 47 shared benchmarks.
Mistral AISummary
- They share 47 benchmarks with published results for both. GPT-4o scores higher in 3 categories and Mistral Large in 6 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-4o leads 52.6 to 40.7.
- The biggest single-benchmark swing is BigCodeBench Complete: 61.1% for GPT-4o and 38.3% for Mistral Large.
- Mistral Large is cheaper at $2 / $6 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- Mistral Large accepts more context: 131K tokens versus 128K.
- Mistral Large has downloadable open weights; the other is API-only.
Side by side
| GPT-4o | Mistral Large | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 28.6 | 31.9 |
| Released | 2024-05-13 | 2024-02-26 |
| Weights | Proprietary | Open |
| Context window | 128K | 131K |
| Max output | 16K | 16K |
| Input $ / M tokens | $2.50 | $2 |
| Output $ / M tokens | $10 | $6 |
| Results tracked | 72 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Large leads
GPT-4o: 24.8 (#328), Mistral Large: 34.3 (#240)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| BigCodeBench Instruct | 51.1% | 30% |
| LiveBench Coding | 51.4% | 47.1% |
| LMArena Coding | 1297 | 1277 |
| BigCodeBench Complete | 61.1% | 38.3% |
| HumanEval+ | 87.2% | 62.2% |
| MBPP+ | 72.2% | 59.5% |
| SWE-bench Verified | 31% | — |
| SWE-bench Verified (bash only) | 21.6% | — |
| Aider Polyglot | 45.3% | — |
| SciCode | — | 36.2% |
| GSO | 0% | — |
| WeirdML | 25.1% | — |
| CadEval | 26% | — |
| ALE-Bench | — | 264.7 |
Agentic & Tool Use Mistral Large leads
GPT-4o: 21.0 (#141), Mistral Large: 28.6 (#89)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 38.4% |
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
Reasoning Mistral Large leads
GPT-4o: 9.4 (#343), Mistral Large: 15.8 (#310)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| SimpleBench | 17.8% | 22.5% |
| CritPt | 0% | 0% |
| LiveBench Reasoning | 55.8% | 43.5% |
| LMArena Hard Prompts | 1281 | 1257 |
| DTBench | 64.5% | 65.1% |
| LiveBench Data Analysis | 60.9% | 50.1% |
| LMCA | 16.6% | 16.7% |
| Epoch Capabilities Index | 128.97 | 128.52 |
| ForecastBench | 57.7 | 57.1 |
| LiveBench | 55.3% | 48.4% |
| ARC-AGI-2 | 0% | — |
| ARC-AGI-1 | 4.5% | — |
| Chess Puzzles | 13% | — |
| EnigmaEval | 0.8% | — |
Math Mistral Large leads
GPT-4o: 10.6 (#312), Mistral Large: 18.2 (#291)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.4% | 8.5% |
| Omni-MATH | 29.3% | 28.1% |
| LiveBench Math | 49.5% | 42.5% |
| LMArena Math | 1285 | 1262 |
| MATH Level 5 | 53.3% | 50.3% |
| FrontierMath (Feb 2025 set) | 0.3% | 0.3% |
| FrontierMath (Tiers 1-3) | 0.4% | — |
Knowledge Mistral Large leads
GPT-4o: 28.8 (#242), Mistral Large: 30.1 (#230)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| GPQA Diamond | 49.2% | 51.3% |
| MMLU-Pro | 71.3% | 59.9% |
| Confabulations | 15.3% | 21.4% |
| Vectara Hallucination Rate | 9.6% | 4.5% |
| GPQA (HELM) | 52% | 43.5% |
| LMArena Expert | 1250 | 1232 |
| MMLU | 88.1% | 80% |
| Humanity's Last Exam | 2.7% | — |
| SimpleQA Verified | 26% | — |
Multimodal Not comparable
GPT-4o: 34.5 (#91), Mistral Large: —
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| LMArena Vision | 1137 | — |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| ScienceQA | 88.5% | — |
Multilingual GPT-4o leads
GPT-4o: 43.2 (#186), Mistral Large: 40.0 (#219)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| LMArena Non-English | 1283 | 1237 |
| LMArena Chinese | 1277 | 1240 |
| LMArena French | 1304 | 1325 |
| LMArena German | 1282 | 1254 |
| LMArena Japanese | 1257 | 1188 |
| LMArena Korean | 1234 | 1202 |
| LMArena Russian | 1286 | 1257 |
| LMArena Spanish | 1292 | 1268 |
Instruction Following Mistral Large leads
GPT-4o: 66.6 (#207), Mistral Large: 67.9 (#191)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| LiveBench Instruction Following | 68.6% | 67.9% |
| IFEval | 81.7% | 87.7% |
| LMArena Instruction Following | 1278 | 1249 |
Long Context GPT-4o leads
GPT-4o: 39.4 (#179), Mistral Large: 38.3 (#199)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| LMArena Longer Query | 1289 | 1261 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference GPT-4o leads
GPT-4o: 52.6 (#166), Mistral Large: 40.7 (#242)
| Benchmark | GPT-4o | Mistral Large |
|---|---|---|
| LMArena Text | 1300 | 1266 |
| LMArena Creative Writing | 1292 | 1243 |
| Short-Story Creative Writing | 81.8% | 69% |
| WildBench | 82.8% | 80.1% |
| LMArena Multi-Turn | 1302 | 1260 |
| LiveBench Language | 47.6% | 39.4% |
| EQ-Bench Creative Writing | — | 985 |
Frequently asked questions
Is GPT-4o better than Mistral Large?
Mistral Large is the stronger model overall, scoring 31.9 to 28.6 on the Noometry Index.
Which is cheaper, GPT-4o or Mistral Large?
Mistral Large is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-4o lists at $2.50 and $10.
Is GPT-4o or Mistral Large better for coding?
Mistral Large scores higher on coding benchmarks: 34.3 versus 24.8 in the Noometry coding category.
Which has the bigger context window?
Mistral Large does, with 131K tokens against 128K.
How many benchmarks do GPT-4o and Mistral Large share?
47 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and Mistral Large has 51.