Model comparison
GPT-3.5-turbo vs Mistral Small
Mistral Small is the stronger model overall, scoring 33.1 to 23.1 on the Noometry Index.
Last verified . 25 shared benchmarks.
Mistral AISummary
- They share 25 benchmarks with published results for both. GPT-3.5-turbo scores higher in 0 categories and Mistral Small in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Mistral Small leads 52.6 to 25.4.
- The biggest single-benchmark swing is MATH Level 5: 15.9% for GPT-3.5-turbo and 46.8% for Mistral Small.
- Mistral Small is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.50 / $1.50 for GPT-3.5-turbo.
- Mistral Small accepts more context: 256K tokens versus 16K.
- Mistral Small has downloadable open weights; the other is API-only.
Side by side
| GPT-3.5-turbo | Mistral Small | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 23.1 | 33.1 |
| Released | 2023-03-01 | 2024-02-26 |
| Weights | Proprietary | Open |
| Context window | 16K | 256K |
| Max output | 4K | 256K |
| Input $ / M tokens | $0.50 | $0.15 |
| Output $ / M tokens | $1.50 | $0.60 |
| Results tracked | 44 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Small leads
GPT-3.5-turbo: 23.5 (#326), Mistral Small: 33.5 (#242)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| BigCodeBench Instruct | 39.1% | 36.1% |
| LMArena Coding | 1137 | 1363 |
| BigCodeBench Complete | 50.6% | 46.6% |
| SciCode | — | 26.5% |
| WeirdML | 3.5% | — |
| LiveBench Coding | — | 36.2% |
| ALE-Bench | — | 497.62 |
| HumanEval+ | 70.7% | — |
| MBPP+ | 69.7% | — |
Agentic & Tool Use Not comparable
GPT-3.5-turbo: —, Mistral Small: 27.9 (#87)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.1% |
| METR Time Horizons | 21.5% | — |
Reasoning Mistral Small leads
GPT-3.5-turbo: 13.3 (#328), Mistral Small: 19.3 (#238)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| LMArena Hard Prompts | 1108 | 1336 |
| DTBench | 48.5% | 70.9% |
| LMCA | 9.7% | 20.6% |
| Kagi LLM Benchmark | — | 37.8% |
| CritPt | — | 0% |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | — | 44.8% |
| Mystery Game Puzzles | 3% | — |
| LiveBench Data Analysis | — | 53.7% |
| Adversarial NLI | 58.1% | — |
| BIG-Bench Hard | 61.6% | — |
| CommonsenseQA 2.0 | 57% | — |
| Epoch Capabilities Index | 118.48 | — |
| ForecastBench | 50.4 | — |
| LiveBench | — | 44% |
| WinoGrande | 81.6% | — |
Math Mistral Small leads
GPT-3.5-turbo: 6.0 (#321), Mistral Small: 15.4 (#287)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.2% | 5.8% |
| LMArena Math | 1142 | 1341 |
| MATH Level 5 | 15.9% | 46.8% |
| FrontierMath (Tiers 1-3) | 0% | — |
| LiveBench Math | — | 39.9% |
| GSM8K | 57.8% | — |
Knowledge Mistral Small leads
GPT-3.5-turbo: 10.0 (#298), Mistral Small: 30.9 (#214)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| GPQA Diamond | 28% | 47.5% |
| LMArena Expert | 1070 | 1293 |
| MMLU | 71.4% | 68.7% |
| Vectara Hallucination Rate | — | 5.1% |
| ARC (AI2) Challenge | 87.4% | — |
| BoolQ | 87% | — |
| OpenBookQA | 86% | — |
| TriviaQA | 85.8% | — |
Multimodal Not comparable
GPT-3.5-turbo: —, Mistral Small: 32.9 (#91)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| LMArena Vision | — | 1142 |
Multilingual Mistral Small leads
GPT-3.5-turbo: 31.6 (#253), Mistral Small: 45.7 (#164)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| LMArena Non-English | 1108 | 1315 |
| LMArena Chinese | 1075 | 1342 |
| LMArena French | 1117 | 1337 |
| LMArena German | 1090 | 1341 |
| LMArena Japanese | 1043 | 1275 |
| LMArena Korean | 1019 | 1260 |
| LMArena Russian | 1123 | 1324 |
| LMArena Spanish | 1120 | 1346 |
Instruction Following Mistral Small leads
GPT-3.5-turbo: 58.1 (#257), Mistral Small: 66.6 (#204)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| LMArena Instruction Following | 1119 | 1310 |
| LiveBench Instruction Following | — | 63.7% |
Long Context Mistral Small leads
GPT-3.5-turbo: 34.0 (#249), Mistral Small: 40.4 (#151)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| LMArena Longer Query | 1121 | 1327 |
Writing & Preference Mistral Small leads
GPT-3.5-turbo: 25.4 (#300), Mistral Small: 52.6 (#165)
| Benchmark | GPT-3.5-turbo | Mistral Small |
|---|---|---|
| LMArena Text | 1125 | 1338 |
| LMArena Creative Writing | 1092 | 1304 |
| LMArena Multi-Turn | 1117 | 1344 |
| EQ-Bench Creative Writing | 451 | — |
| LiveBench Language | — | 30.5% |
Frequently asked questions
Is GPT-3.5-turbo better than Mistral Small?
Mistral Small is the stronger model overall, scoring 33.1 to 23.1 on the Noometry Index.
Which is cheaper, GPT-3.5-turbo or Mistral Small?
Mistral Small is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-3.5-turbo lists at $0.50 and $1.50.
Is GPT-3.5-turbo or Mistral Small better for coding?
Mistral Small scores higher on coding benchmarks: 33.5 versus 23.5 in the Noometry coding category.
Which has the bigger context window?
Mistral Small does, with 256K tokens against 16K.
How many benchmarks do GPT-3.5-turbo and Mistral Small share?
25 benchmarks have published results for both models. GPT-3.5-turbo has 44 scored results on Noometry and Mistral Small has 39.