Model comparison
Mistral Small 3 vs Qwen2.5-Coder-32B
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 31.2 on the Noometry Index. Mistral Small 3 costs 13× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Last verified . 15 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 15 benchmarks with published results for both. Mistral Small 3 scores higher in 2 categories and Qwen2.5-Coder-32B in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen2.5-Coder-32B leads 33.3 to 16.3.
- The biggest single-benchmark swing is BigCodeBench Complete: 50.4% for Mistral Small 3 and 58% for Qwen2.5-Coder-32B.
- Mistral Small 3 is cheaper at $0.05 / $0.08 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
Side by side
| Mistral Small 3 | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 31.2 | 33.4 |
| Released | 2025-01-30 | 2024-09-18 |
| Weights | Open | Open |
| Context window | 33K | 33K |
| Max output | 16K | 29K |
| Input $ / M tokens | $0.05 | $0.66 |
| Output $ / M tokens | $0.08 | $1 |
| Results tracked | 24 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Small 3 leads
Mistral Small 3: 36.5 (#207), Qwen2.5-Coder-32B: 22.6 (#333)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| BigCodeBench Instruct | 45.3% | 49% |
| LMArena Coding | 1246 | 1276 |
| BigCodeBench Complete | 50.4% | 58% |
| SWE-bench Verified (bash only) | — | 9% |
| Aider Polyglot | — | 16.4% |
| LiveBench Coding | — | 56.9% |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Reasoning Qwen2.5-Coder-32B leads
Mistral Small 3: 18.9 (#273), Qwen2.5-Coder-32B: 21.2 (#225)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Hard Prompts | 1233 | 1251 |
| Epoch Capabilities Index | 127.07 | 119.49 |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | — | 42.1% |
| LiveBench Data Analysis | — | 49.9% |
| HellaSwag | — | 83% |
| LiveBench | — | 46.2% |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
Mistral Small 3: 16.3 (#295), Qwen2.5-Coder-32B: 33.3 (#204)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Math | 1240 | 1251 |
| OTIS Mock AIME 2024-2025 | 6.7% | — |
| LiveBench Math | — | 46.6% |
| GSM8K | — | 93% |
Knowledge Qwen2.5-Coder-32B leads
Mistral Small 3: 25.1 (#263), Qwen2.5-Coder-32B: 33.4 (#203)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1202 | 1221 |
| GPQA Diamond | 47.3% | — |
| Confabulations | 25.2% | — |
| ARC (AI2) Challenge | — | 70.5% |
| MMLU | — | 79.1% |
Multilingual Too close to call
Mistral Small 3: 37.3 (#236), Qwen2.5-Coder-32B: 37.8 (#235)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1198 | 1205 |
| LMArena Chinese | 1204 | 1222 |
| LMArena Russian | 1216 | 1228 |
| LMArena French | 1203 | — |
| LMArena German | 1211 | — |
| LMArena Japanese | 1111 | — |
| LMArena Korean | 1188 | — |
Instruction Following Mistral Small 3 leads
Mistral Small 3: 63.7 (#229), Qwen2.5-Coder-32B: 61.4 (#245)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Instruction Following | 1214 | 1223 |
| LiveBench Instruction Following | — | 58.7% |
Long Context Too close to call
Mistral Small 3: 37.8 (#211), Qwen2.5-Coder-32B: 38.0 (#208)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1246 | 1251 |
Writing & Preference Qwen2.5-Coder-32B leads
Mistral Small 3: 32.2 (#280), Qwen2.5-Coder-32B: 41.6 (#240)
| Benchmark | Mistral Small 3 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1234 | 1230 |
| LMArena Creative Writing | 1195 | 1174 |
| LMArena Multi-Turn | 1217 | 1222 |
| EQ-Bench Creative Writing | 707 | — |
| LiveBench Language | — | 23.3% |
Frequently asked questions
Is Mistral Small 3 better than Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 31.2 on the Noometry Index. Mistral Small 3 costs 13× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Which is cheaper, Mistral Small 3 or Qwen2.5-Coder-32B?
Mistral Small 3 is cheaper. It lists at $0.05 per million input tokens and $0.08 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is Mistral Small 3 or Qwen2.5-Coder-32B better for coding?
Mistral Small 3 scores higher on coding benchmarks: 36.5 versus 22.6 in the Noometry coding category.
Which has the bigger context window?
Both accept 33K tokens.
How many benchmarks do Mistral Small 3 and Qwen2.5-Coder-32B share?
15 benchmarks have published results for both models. Mistral Small 3 has 24 scored results on Noometry and Qwen2.5-Coder-32B has 31.