Model comparison
MiMo-V2-Pro vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 43.0 on the Noometry Index. MiMo-V2-Pro costs 5.5× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. MiMo-V2-Pro scores higher in 0 categories and Qwen3.8 Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 39.5.
- The biggest single-benchmark swing is NYT Connections (extended): 25.8% for MiMo-V2-Pro and 88.3% for Qwen3.8 Max.
- MiMo-V2-Pro is cheaper at $0.43 / $0.87 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- MiMo-V2-Pro accepts more context: 1.05M tokens versus 1M.
Side by side
| MiMo-V2-Pro | Qwen3.8 Max | |
|---|---|---|
| Provider | Xiaomi | Alibaba (Qwen) |
| Noometry Index | 43.0 | 56.8 |
| Released | 2026-03-18 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $0.43 | $2 |
| Output $ / M tokens | $0.87 | $6 |
| Results tracked | 23 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
MiMo-V2-Pro: 43.8 (#83), Qwen3.8 Max: 53.5 (#29)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena WebDev | 1433 | 1674 |
| LMArena Coding | 1476 | 1502 |
| DeepSWE | — | 57.5% |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| ALE-Bench | 785.17 | — |
Agentic & Tool Use Not comparable
MiMo-V2-Pro: —, Qwen3.8 Max: 45.4 (#14)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
MiMo-V2-Pro: 22.1 (#206), Qwen3.8 Max: 54.4 (#26)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 25.8% | 88.3% |
| LMArena Hard Prompts | 1457 | 1496 |
| CritPt | — | 20% |
| Chess Puzzles | — | 40% |
| Thematic Generalization | 45.9% | — |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LMCA | — | 46.2% |
| Epoch Capabilities Index | — | 156.41 |
Math Qwen3.8 Max leads
MiMo-V2-Pro: 39.5 (#102), Qwen3.8 Max: 73.2 (#20)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Math | 1447 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |
Knowledge Qwen3.8 Max leads
MiMo-V2-Pro: 41.4 (#111), Qwen3.8 Max: 61.7 (#27)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1478 | 1507 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |
Multimodal Not comparable
MiMo-V2-Pro: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
MiMo-V2-Pro: 52.7 (#81), Qwen3.8 Max: 56.7 (#18)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1416 | 1472 |
| LMArena Chinese | 1456 | 1538 |
| LMArena French | 1469 | 1503 |
| LMArena German | 1417 | 1483 |
| LMArena Japanese | 1366 | 1467 |
| LMArena Korean | 1400 | 1461 |
| LMArena Russian | 1427 | 1481 |
| LMArena Spanish | 1457 | 1492 |
Instruction Following Qwen3.8 Max leads
MiMo-V2-Pro: 76.0 (#49), Qwen3.8 Max: 77.6 (#17)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1445 | 1479 |
Long Context Qwen3.8 Max leads
MiMo-V2-Pro: 41.5 (#138), Qwen3.8 Max: 45.6 (#31)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1455 | 1489 |
| CL-bench | 15.7% | — |
| CL-bench Life | 6.9% | — |
Writing & Preference Qwen3.8 Max leads
MiMo-V2-Pro: 62.8 (#70), Qwen3.8 Max: 67.1 (#30)
| Benchmark | MiMo-V2-Pro | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1436 | 1483 |
| LMArena Creative Writing | 1415 | 1479 |
| LMArena Multi-Turn | 1456 | 1489 |
Frequently asked questions
Is MiMo-V2-Pro better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 43.0 on the Noometry Index. MiMo-V2-Pro costs 5.5× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, MiMo-V2-Pro or Qwen3.8 Max?
MiMo-V2-Pro is cheaper. It lists at $0.43 per million input tokens and $0.87 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is MiMo-V2-Pro or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 43.8 in the Noometry coding category.
Which has the bigger context window?
MiMo-V2-Pro does, with 1.05M tokens against 1M.
How many benchmarks do MiMo-V2-Pro and Qwen3.8 Max share?
19 benchmarks have published results for both models. MiMo-V2-Pro has 23 scored results on Noometry and Qwen3.8 Max has 39.