Model comparison
GPT-4o vs Mistral Small
Mistral Small is the stronger model overall, scoring 33.4 to 28.6 on the Noometry Index.
Last verified . 35 shared benchmarks.
Mistral AISummary
- They share 35 benchmarks with published results for both. GPT-4o scores higher in 3 categories and Mistral Small in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Mistral Small leads 19.8 to 9.4.
- The biggest single-benchmark swing is LiveBench Language: 47.6% for GPT-4o and 30.5% for Mistral Small.
- Mistral Small is cheaper at $0.15 / $0.60 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- Mistral Small accepts more context: 262K tokens versus 128K.
- Mistral Small has downloadable open weights; the other is API-only.
Side by side
| GPT-4o | Mistral Small | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 28.6 | 33.4 |
| Released | 2024-05-13 | 2024-02-26 |
| Weights | Proprietary | Open |
| Context window | 128K | 262K |
| Max output | 16K | 256K |
| Input $ / M tokens | $2.50 | $0.15 |
| Output $ / M tokens | $10 | $0.60 |
| Results tracked | 72 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Small leads
GPT-4o: 24.8 (#328), Mistral Small: 34.0 (#247)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| BigCodeBench Instruct | 51.1% | 36.1% |
| LiveBench Coding | 51.4% | 36.2% |
| LMArena Coding | 1297 | 1362 |
| BigCodeBench Complete | 61.1% | 46.6% |
| SWE-bench Verified | 31% | — |
| SWE-bench Verified (bash only) | 21.6% | — |
| Aider Polyglot | 45.3% | — |
| SciCode | — | 26.5% |
| GSO | 0% | — |
| WeirdML | 25.1% | — |
| CadEval | 26% | — |
| ALE-Bench | — | 497.62 |
| HumanEval+ | 87.2% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Mistral Small leads
GPT-4o: 21.0 (#141), Mistral Small: 28.1 (#93)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.1% |
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
Reasoning Mistral Small leads
GPT-4o: 9.4 (#343), Mistral Small: 19.8 (#250)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| CritPt | 0% | 0% |
| LiveBench Reasoning | 55.8% | 44.8% |
| LMArena Hard Prompts | 1281 | 1335 |
| DTBench | 64.5% | 70.9% |
| LiveBench Data Analysis | 60.9% | 53.7% |
| LMCA | 16.6% | 20.6% |
| LiveBench | 55.3% | 44% |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 17.8% | — |
| Kagi LLM Benchmark | — | 37.8% |
| ARC-AGI-1 | 4.5% | — |
| Chess Puzzles | 13% | — |
| EnigmaEval | 0.8% | — |
| Epoch Capabilities Index | 128.97 | — |
| ForecastBench | 57.7 | — |
Math Mistral Small leads
GPT-4o: 10.6 (#312), Mistral Small: 16.4 (#293)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.4% | 5.8% |
| LiveBench Math | 49.5% | 39.9% |
| LMArena Math | 1285 | 1341 |
| MATH Level 5 | 53.3% | 46.8% |
| FrontierMath (Tiers 1-3) | 0.4% | — |
| Omni-MATH | 29.3% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Mistral Small leads
GPT-4o: 28.8 (#242), Mistral Small: 31.0 (#222)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| GPQA Diamond | 49.2% | 47.5% |
| Vectara Hallucination Rate | 9.6% | 5.1% |
| LMArena Expert | 1250 | 1291 |
| MMLU | 88.1% | 68.7% |
| Humanity's Last Exam | 2.7% | — |
| SimpleQA Verified | 26% | — |
| MMLU-Pro | 71.3% | — |
| Confabulations | 15.3% | — |
| GPQA (HELM) | 52% | — |
Multimodal GPT-4o leads
GPT-4o: 34.5 (#91), Mistral Small: 33.5 (#96)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| LMArena Vision | 1137 | 1142 |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| ScienceQA | 88.5% | — |
Multilingual Mistral Small leads
GPT-4o: 43.2 (#186), Mistral Small: 45.5 (#169)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| LMArena Non-English | 1283 | 1315 |
| LMArena Chinese | 1277 | 1340 |
| LMArena French | 1304 | 1337 |
| LMArena German | 1282 | 1340 |
| LMArena Japanese | 1257 | 1275 |
| LMArena Korean | 1234 | 1259 |
| LMArena Russian | 1286 | 1324 |
| LMArena Spanish | 1292 | 1346 |
Instruction Following Too close to call
GPT-4o: 66.6 (#207), Mistral Small: 66.4 (#209)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| LiveBench Instruction Following | 68.6% | 63.7% |
| LMArena Instruction Following | 1278 | 1310 |
| IFEval | 81.7% | — |
Long Context Too close to call
GPT-4o: 39.4 (#179), Mistral Small: 40.4 (#156)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| LMArena Longer Query | 1289 | 1327 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Too close to call
GPT-4o: 52.6 (#166), Mistral Small: 52.5 (#171)
| Benchmark | GPT-4o | Mistral Small |
|---|---|---|
| LMArena Text | 1300 | 1338 |
| LMArena Creative Writing | 1292 | 1305 |
| LMArena Multi-Turn | 1302 | 1344 |
| LiveBench Language | 47.6% | 30.5% |
| Short-Story Creative Writing | 81.8% | — |
| WildBench | 82.8% | — |
Frequently asked questions
Is GPT-4o better than Mistral Small?
Mistral Small is the stronger model overall, scoring 33.4 to 28.6 on the Noometry Index.
Which is cheaper, GPT-4o or Mistral Small?
Mistral Small is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-4o lists at $2.50 and $10.
Is GPT-4o or Mistral Small better for coding?
Mistral Small scores higher on coding benchmarks: 34.0 versus 24.8 in the Noometry coding category.
Which has the bigger context window?
Mistral Small does, with 262K tokens against 128K.
How many benchmarks do GPT-4o and Mistral Small share?
35 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and Mistral Small has 39.