Model comparison
MiniMax-M2.7 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 37.7 on the Noometry Index. MiniMax-M2.7 costs 5.7× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 23 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. MiniMax-M2.7 scores higher in 0 categories and Qwen3.8 Max in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 25.9.
- The biggest single-benchmark swing is NYT Connections (extended): 24.7% for MiniMax-M2.7 and 88.3% for Qwen3.8 Max.
- MiniMax-M2.7 is cheaper at $0.30 / $1.20 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Qwen3.8 Max accepts more context: 1M tokens versus 205K.
- MiniMax-M2.7 has downloadable open weights; the other is API-only.
Side by side
| MiniMax-M2.7 | Qwen3.8 Max | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 37.7 | 56.8 |
| Released | 2026-03-18 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | 205K | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $0.30 | $2 |
| Output $ / M tokens | $1.20 | $6 |
| Results tracked | 30 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
MiniMax-M2.7: 41.8 (#120), Qwen3.8 Max: 53.5 (#29)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena WebDev | 1398 | 1674 |
| SciCode | 47% | 53.2% |
| LMArena Coding | 1454 | 1502 |
| DeepSWE | — | 57.5% |
| FrontierSWE | — | 17.8% |
| WeirdML | 37% | — |
| ALE-Bench | 599.25 | — |
Agentic & Tool Use Qwen3.8 Max leads
MiniMax-M2.7: 25.1 (#111), Qwen3.8 Max: 45.4 (#14)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| Terminal-Bench | 45.1% | — |
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| ExploitBench | 13.3% | — |
| GBAEval | 0% | — |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
MiniMax-M2.7: 19.7 (#253), Qwen3.8 Max: 54.4 (#26)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 24.7% | 88.3% |
| CritPt | 0.6% | 20% |
| LMArena Hard Prompts | 1422 | 1496 |
| Epoch Capabilities Index | 145.85 | 156.41 |
| Chess Puzzles | — | 40% |
| Thematic Generalization | 39.3% | — |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LMCA | — | 46.2% |
Math Qwen3.8 Max leads
MiniMax-M2.7: 25.9 (#263), Qwen3.8 Max: 73.2 (#20)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| ProofBench | 3% | 58% |
| LMArena Math | 1420 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
Knowledge Qwen3.8 Max leads
MiniMax-M2.7: 37.7 (#152), Qwen3.8 Max: 61.7 (#27)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1444 | 1507 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |
| Vectara Hallucination Rate | 12.9% | — |
Multimodal Not comparable
MiniMax-M2.7: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
MiniMax-M2.7: 50.3 (#123), Qwen3.8 Max: 56.7 (#18)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1382 | 1472 |
| LMArena Chinese | 1441 | 1538 |
| LMArena French | 1421 | 1503 |
| LMArena German | 1398 | 1483 |
| LMArena Japanese | 1262 | 1467 |
| LMArena Korean | 1313 | 1461 |
| LMArena Russian | 1383 | 1481 |
| LMArena Spanish | 1403 | 1492 |
Instruction Following Qwen3.8 Max leads
MiniMax-M2.7: 74.1 (#103), Qwen3.8 Max: 77.6 (#17)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1405 | 1479 |
Long Context Qwen3.8 Max leads
MiniMax-M2.7: 43.3 (#99), Qwen3.8 Max: 45.6 (#31)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1419 | 1489 |
Writing & Preference Qwen3.8 Max leads
MiniMax-M2.7: 58.9 (#112), Qwen3.8 Max: 67.1 (#30)
| Benchmark | MiniMax-M2.7 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1405 | 1483 |
| LMArena Creative Writing | 1354 | 1479 |
| LMArena Multi-Turn | 1412 | 1489 |
Frequently asked questions
Is MiniMax-M2.7 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 37.7 on the Noometry Index. MiniMax-M2.7 costs 5.7× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M2.7 or Qwen3.8 Max?
MiniMax-M2.7 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is MiniMax-M2.7 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 41.8 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 205K.
How many benchmarks do MiniMax-M2.7 and Qwen3.8 Max share?
23 benchmarks have published results for both models. MiniMax-M2.7 has 30 scored results on Noometry and Qwen3.8 Max has 39.