Model comparison
MiniMax-M2.5 vs Qwen2-72B
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 30.0 on the Noometry Index.
Last verified . 18 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. MiniMax-M2.5 scores higher in 7 categories and Qwen2-72B in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where MiniMax-M2.5 leads 48.1 to 29.1.
Side by side
| MiniMax-M2.5 | Qwen2-72B | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 38.3 | 30.0 |
| Released | 2026-02-12 | 2024-06-07 |
| Weights | Open | Open |
| Context window | 205K | — |
| Max output | 131K | — |
| Input $ / M tokens | $0.30 | — |
| Output $ / M tokens | $1.20 | — |
| Results tracked | 33 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
MiniMax-M2.5: 48.1 (#58), Qwen2-72B: 29.1 (#310)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Coding | 1381 | 1196 |
| SWE-bench Verified (bash only) | 75.8% | — |
| LMArena WebDev | 1387 | — |
| SWE-bench Multilingual | 68.3% | — |
| WeirdML | — | 11.3% |
| BigCodeBench Instruct | — | 38.5% |
| BigCodeBench Complete | — | 54% |
| ALE-Bench | 618.17 | — |
Agentic & Tool Use MiniMax-M2.5 leads
MiniMax-M2.5: 30.4 (#77), Qwen2-72B: 17.0 (#146)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| Terminal-Bench | 42.7% | — |
| TheAgentCompany | — | 1.1% |
| METR Time Horizons | — | 29.9% |
| Vending-Bench 2 | -23.16 | — |
Reasoning Qwen2-72B leads
MiniMax-M2.5: 17.5 (#292), Qwen2-72B: 23.2 (#181)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1372 | 1191 |
| Epoch Capabilities Index | 146.68 | 125.28 |
| ARC-AGI-2 | 4.9% | — |
| Kagi LLM Benchmark | 55.2% | — |
| NYT Connections (extended) | 16.8% | — |
| ARC-AGI-1 | 63.7% | — |
Math Qwen2-72B leads
MiniMax-M2.5: 26.9 (#253), Qwen2-72B: 30.2 (#236)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Math | 1378 | 1235 |
| ProofBench | 4% | — |
| MATH Level 5 | — | 39.1% |
Knowledge MiniMax-M2.5 leads
MiniMax-M2.5: 39.2 (#135), Qwen2-72B: 21.2 (#275)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Expert | 1379 | 1171 |
| GPQA Diamond | — | 40.8% |
| Vectara Hallucination Rate | 9.1% | — |
| MMLU | — | 82.4% |
Multilingual MiniMax-M2.5 leads
MiniMax-M2.5: 47.1 (#152), Qwen2-72B: 35.9 (#244)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1338 | 1176 |
| LMArena Chinese | 1393 | 1240 |
| LMArena French | 1362 | 1170 |
| LMArena German | 1362 | 1151 |
| LMArena Japanese | 1171 | 1111 |
| LMArena Korean | 1232 | 1083 |
| LMArena Russian | 1358 | 1169 |
| LMArena Spanish | 1354 | 1169 |
Instruction Following MiniMax-M2.5 leads
MiniMax-M2.5: 71.5 (#148), Qwen2-72B: 61.7 (#241)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1353 | 1181 |
Long Context MiniMax-M2.5 leads
MiniMax-M2.5: 37.5 (#216), Qwen2-72B: 36.1 (#235)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1366 | 1192 |
| CL-bench | 11.4% | — |
| CL-bench Life | 6.3% | — |
Writing & Preference MiniMax-M2.5 leads
MiniMax-M2.5: 53.9 (#153), Qwen2-72B: 40.8 (#241)
| Benchmark | MiniMax-M2.5 | Qwen2-72B |
|---|---|---|
| LMArena Text | 1359 | 1203 |
| LMArena Creative Writing | 1331 | 1181 |
| LMArena Multi-Turn | 1364 | 1196 |
| EQ-Bench Creative Writing | 1361 | — |
Frequently asked questions
Is MiniMax-M2.5 better than Qwen2-72B?
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 30.0 on the Noometry Index.
Is MiniMax-M2.5 or Qwen2-72B better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 29.1 in the Noometry coding category.
How many benchmarks do MiniMax-M2.5 and Qwen2-72B share?
18 benchmarks have published results for both models. MiniMax-M2.5 has 33 scored results on Noometry and Qwen2-72B has 26.