Model comparison
MiniMax-M2.5 vs Qwen2.5 72B Instruct
MiniMax-M2.5 is the stronger model overall, scoring 38.1 to 31.5 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 Qwen2.5 72B Instruct in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where MiniMax-M2.5 leads 47.9 to 32.8.
- MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- MiniMax-M2.5 accepts more context: 205K tokens versus 131K.
Side by side
| MiniMax-M2.5 | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 38.1 | 31.5 |
| Released | 2026-02-12 | 2024-09 |
| Weights | Open | Open |
| Context window | 205K | 131K |
| Max output | 131K | 8K |
| Input $ / M tokens | $0.30 | $1.40 |
| Output $ / M tokens | $1.20 | $5.60 |
| Results tracked | 33 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
MiniMax-M2.5: 47.9 (#53), Qwen2.5 72B Instruct: 32.8 (#256)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Coding | 1380 | 1293 |
| SWE-bench Verified (bash only) | 75.8% | — |
| LMArena WebDev | 1386 | — |
| SWE-bench Multilingual | 68.3% | — |
| WeirdML | — | 16% |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
| ALE-Bench | 618.17 | — |
Agentic & Tool Use MiniMax-M2.5 leads
MiniMax-M2.5: 30.2 (#70), Qwen2.5 72B Instruct: 22.0 (#124)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| Terminal-Bench | 42.7% | — |
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| METR Time Horizons | — | 35.8% |
| Vending-Bench 2 | -23.16 | — |
Reasoning Qwen2.5 72B Instruct leads
MiniMax-M2.5: 17.6 (#281), Qwen2.5 72B Instruct: 21.4 (#197)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1372 | 1271 |
| Epoch Capabilities Index | 146.68 | 129.01 |
| ARC-AGI-2 | 4.9% | — |
| Kagi LLM Benchmark | 55.2% | — |
| NYT Connections (extended) | 16.8% | — |
| ARC-AGI-1 | 63.7% | — |
| DTBench | — | 62.9% |
| LMCA | — | 13.4% |
| BIG-Bench Hard | — | 79.8% |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math MiniMax-M2.5 leads
MiniMax-M2.5: 25.8 (#242), Qwen2.5 72B Instruct: 18.1 (#281)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Math | 1378 | 1283 |
| OTIS Mock AIME 2024-2025 | — | 8.1% |
| ProofBench | 4% | — |
| Omni-MATH | — | 33% |
| MATH Level 5 | — | 63.2% |
Knowledge MiniMax-M2.5 leads
MiniMax-M2.5: 38.6 (#132), Qwen2.5 72B Instruct: 26.9 (#247)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Expert | 1378 | 1245 |
| GPQA Diamond | — | 49.1% |
| MMLU-Pro | — | 63.1% |
| Confabulations | — | 19.1% |
| Vectara Hallucination Rate | 9.1% | — |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multilingual MiniMax-M2.5 leads
MiniMax-M2.5: 47.3 (#147), Qwen2.5 72B Instruct: 41.2 (#208)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1338 | 1252 |
| LMArena Chinese | 1392 | 1272 |
| LMArena French | 1363 | 1279 |
| LMArena German | 1361 | 1234 |
| LMArena Japanese | 1169 | 1180 |
| LMArena Korean | 1232 | 1188 |
| LMArena Russian | 1359 | 1264 |
| LMArena Spanish | 1354 | 1255 |
Instruction Following MiniMax-M2.5 leads
MiniMax-M2.5: 71.7 (#143), Qwen2.5 72B Instruct: 65.7 (#216)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1353 | 1254 |
| IFEval | — | 80.6% |
Long Context Qwen2.5 72B Instruct leads
MiniMax-M2.5: 37.4 (#211), Qwen2.5 72B Instruct: 38.9 (#183)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1367 | 1282 |
| CL-bench | 11.4% | — |
| CL-bench Life | 6.3% | — |
Writing & Preference MiniMax-M2.5 leads
MiniMax-M2.5: 54.1 (#149), Qwen2.5 72B Instruct: 46.9 (#210)
| Benchmark | MiniMax-M2.5 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1359 | 1269 |
| LMArena Creative Writing | 1331 | 1221 |
| LMArena Multi-Turn | 1364 | 1272 |
| EQ-Bench Creative Writing | 1361 | — |
| WildBench | — | 80.2% |
Frequently asked questions
Is MiniMax-M2.5 better than Qwen2.5 72B Instruct?
MiniMax-M2.5 is the stronger model overall, scoring 38.1 to 31.5 on the Noometry Index.
Which is cheaper, MiniMax-M2.5 or Qwen2.5 72B Instruct?
MiniMax-M2.5 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is MiniMax-M2.5 or Qwen2.5 72B Instruct better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 47.9 versus 32.8 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2.5 does, with 205K tokens against 131K.
How many benchmarks do MiniMax-M2.5 and Qwen2.5 72B Instruct share?
18 benchmarks have published results for both models. MiniMax-M2.5 has 33 scored results on Noometry and Qwen2.5 72B Instruct has 43.