Model comparison
MiniMax-M2.5 vs Qwen3.6 Max Preview
Qwen3.6 Max Preview is the stronger model overall, scoring 51.5 to 38.3 on the Noometry Index. MiniMax-M2.5 costs 5.6× less per token, which makes it the better buy when Qwen3.6 Max Preview's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
MiniMax
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. MiniMax-M2.5 scores higher in 0 categories and Qwen3.6 Max Preview in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.6 Max Preview leads 54.1 to 26.9.
- The biggest single-benchmark swing is NYT Connections (extended): 16.8% for MiniMax-M2.5 and 74.1% for Qwen3.6 Max Preview.
- MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.30 / $7.80 for Qwen3.6 Max Preview.
- Qwen3.6 Max Preview accepts more context: 262K tokens versus 205K.
- MiniMax-M2.5 has downloadable open weights; the other is API-only.
Side by side
| MiniMax-M2.5 | Qwen3.6 Max Preview | |
|---|---|---|
| Provider | MiniMax | Alibaba (Qwen) |
| Noometry Index | 38.3 | 51.5 |
| Released | 2026-02-12 | 2026-04-20 |
| Weights | Open | Proprietary |
| Context window | 205K | 262K |
| Max output | 131K | 66K |
| Input $ / M tokens | $0.30 | $1.30 |
| Output $ / M tokens | $1.20 | $7.80 |
| Results tracked | 33 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
MiniMax-M2.5: 48.1 (#58), Qwen3.6 Max Preview: 48.7 (#54)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena WebDev | 1387 | 1482 |
| LMArena Coding | 1381 | 1471 |
| SWE-bench Verified | — | 76.7% |
| SWE-bench Verified (bash only) | 75.8% | — |
| SWE-bench Multilingual | 68.3% | — |
| ALE-Bench | 618.17 | — |
Agentic & Tool Use Not comparable
MiniMax-M2.5: 30.4 (#77), Qwen3.6 Max Preview: —
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| Vending-Bench 2 | -23.16 | 4,254 |
| Terminal-Bench | 42.7% | — |
Reasoning Qwen3.6 Max Preview leads
MiniMax-M2.5: 17.5 (#292), Qwen3.6 Max Preview: 41.7 (#53)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| NYT Connections (extended) | 16.8% | 74.1% |
| LMArena Hard Prompts | 1372 | 1457 |
| Epoch Capabilities Index | 146.68 | 149.24 |
| ARC-AGI-2 | 4.9% | — |
| SimpleBench | — | 63% |
| Kagi LLM Benchmark | 55.2% | — |
| ARC-AGI-1 | 63.7% | — |
| Chess Puzzles | — | 20% |
| Mystery Game Puzzles | — | 19% |
| DTBench | — | 87.2% |
| LMCA | — | 42.5% |
Math Qwen3.6 Max Preview leads
MiniMax-M2.5: 26.9 (#253), Qwen3.6 Max Preview: 54.1 (#46)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Math | 1378 | 1465 |
| OTIS Mock AIME 2024-2025 | — | 91.1% |
| ProofBench | 4% | — |
| FrontierMath (Feb 2025 set) | — | 23.1% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge Qwen3.6 Max Preview leads
MiniMax-M2.5: 39.2 (#135), Qwen3.6 Max Preview: 57.6 (#39)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Expert | 1379 | 1478 |
| GPQA Diamond | — | 87.4% |
| SimpleQA Verified | — | 52% |
| Vectara Hallucination Rate | 9.1% | — |
Multilingual Qwen3.6 Max Preview leads
MiniMax-M2.5: 47.1 (#152), Qwen3.6 Max Preview: 54.2 (#48)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Non-English | 1338 | 1437 |
| LMArena Chinese | 1393 | 1487 |
| LMArena French | 1362 | 1449 |
| LMArena Russian | 1358 | 1445 |
| LMArena Spanish | 1354 | 1454 |
| LMArena German | 1362 | — |
| LMArena Japanese | 1171 | — |
| LMArena Korean | 1232 | — |
Instruction Following Qwen3.6 Max Preview leads
MiniMax-M2.5: 71.5 (#148), Qwen3.6 Max Preview: 75.7 (#55)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Instruction Following | 1353 | 1438 |
Long Context Qwen3.6 Max Preview leads
MiniMax-M2.5: 37.5 (#216), Qwen3.6 Max Preview: 44.6 (#61)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Longer Query | 1366 | 1457 |
| CL-bench | 11.4% | — |
| CL-bench Life | 6.3% | — |
Writing & Preference Qwen3.6 Max Preview leads
MiniMax-M2.5: 53.9 (#153), Qwen3.6 Max Preview: 63.8 (#60)
| Benchmark | MiniMax-M2.5 | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Text | 1359 | 1447 |
| LMArena Creative Writing | 1331 | 1435 |
| LMArena Multi-Turn | 1364 | 1456 |
| EQ-Bench Creative Writing | 1361 | — |
Frequently asked questions
Is MiniMax-M2.5 better than Qwen3.6 Max Preview?
Qwen3.6 Max Preview is the stronger model overall, scoring 51.5 to 38.3 on the Noometry Index. MiniMax-M2.5 costs 5.6× less per token, which makes it the better buy when Qwen3.6 Max Preview's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M2.5 or Qwen3.6 Max Preview?
MiniMax-M2.5 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Qwen3.6 Max Preview lists at $1.30 and $7.80.
Is MiniMax-M2.5 or Qwen3.6 Max Preview better for coding?
They score almost the same on coding (48.1 vs 48.7); test both on your own repository before choosing.
Which has the bigger context window?
Qwen3.6 Max Preview does, with 262K tokens against 205K.
How many benchmarks do MiniMax-M2.5 and Qwen3.6 Max Preview share?
18 benchmarks have published results for both models. MiniMax-M2.5 has 33 scored results on Noometry and Qwen3.6 Max Preview has 29.