Model comparison
Ministral 8B vs Qwen2-72B
Qwen2-72B is the stronger model overall, scoring 29.5 to 27.8 on the Noometry Index.
Last verified . 14 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 14 benchmarks with published results for both. Ministral 8B scores higher in 2 categories and Qwen2-72B in 7 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen2-72B leads 21.2 to 12.5.
- The biggest single-benchmark swing is MATH Level 5: 14.9% for Ministral 8B and 39.1% for Qwen2-72B.
Side by side
| Ministral 8B | Qwen2-72B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 27.8 | 29.5 |
| Released | 2024-10-01 | 2024-06-07 |
| Weights | Open | Open |
| Context window | 128K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.10 | — |
| Output $ / M tokens | $0.10 | — |
| Results tracked | 17 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Ministral 8B leads
Ministral 8B: 34.4 (#226), Qwen2-72B: 28.7 (#303)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Coding | 1202 | 1196 |
| WeirdML | — | 11.3% |
| BigCodeBench Instruct | — | 38.5% |
| BigCodeBench Complete | — | 54% |
Agentic & Tool Use Too close to call
Ministral 8B: 16.2 (#140), Qwen2-72B: 16.9 (#138)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 11.1% | — |
| TheAgentCompany | — | 1.1% |
| METR Time Horizons | — | 29.9% |
Reasoning Qwen2-72B leads
Ministral 8B: 17.7 (#279), Qwen2-72B: 22.2 (#184)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1191 |
| DTBench | 45.7% | — |
| Epoch Capabilities Index | — | 125.27 |
Math Qwen2-72B leads
Ministral 8B: 23.7 (#262), Qwen2-72B: 27.9 (#232)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Math | 1188 | 1235 |
| MATH Level 5 | 14.9% | 39.1% |
Knowledge Qwen2-72B leads
Ministral 8B: 12.5 (#292), Qwen2-72B: 21.2 (#270)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| GPQA Diamond | 27.1% | 40.8% |
| LMArena Expert | 1170 | 1171 |
| Vectara Hallucination Rate | 7.4% | — |
| MMLU | — | 82.4% |
Multilingual Too close to call
Ministral 8B: 35.3 (#242), Qwen2-72B: 36.0 (#239)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1165 | 1176 |
| LMArena Chinese | 1193 | 1240 |
| LMArena Russian | 1195 | 1169 |
| LMArena French | — | 1169 |
| LMArena German | — | 1151 |
| LMArena Japanese | — | 1111 |
| LMArena Korean | — | 1083 |
| LMArena Spanish | — | 1168 |
Instruction Following Qwen2-72B leads
Ministral 8B: 60.7 (#245), Qwen2-72B: 62.0 (#236)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1161 | 1181 |
Long Context Too close to call
Ministral 8B: 36.7 (#222), Qwen2-72B: 36.1 (#230)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1212 | 1193 |
Writing & Preference Qwen2-72B leads
Ministral 8B: 39.8 (#241), Qwen2-72B: 41.0 (#236)
| Benchmark | Ministral 8B | Qwen2-72B |
|---|---|---|
| LMArena Text | 1191 | 1203 |
| LMArena Creative Writing | 1175 | 1181 |
| LMArena Multi-Turn | 1166 | 1195 |
Frequently asked questions
Is Ministral 8B better than Qwen2-72B?
Qwen2-72B is the stronger model overall, scoring 29.5 to 27.8 on the Noometry Index.
Is Ministral 8B or Qwen2-72B better for coding?
Ministral 8B scores higher on coding benchmarks: 34.4 versus 28.7 in the Noometry coding category.
How many benchmarks do Ministral 8B and Qwen2-72B share?
14 benchmarks have published results for both models. Ministral 8B has 17 scored results on Noometry and Qwen2-72B has 26.