Model comparison
MiMo-V2.5 vs Qwen3-4B
MiMo-V2.5 is the stronger model overall, scoring 43.4 to 31.9 on the Noometry Index.
Last verified . 0 shared benchmarks.
Alibaba (Qwen)Summary
- The widest gap is in reasoning, where MiMo-V2.5 leads 28.6 to 19.2.
Side by side
| MiMo-V2.5 | Qwen3-4B | |
|---|---|---|
| Provider | Xiaomi | Alibaba (Qwen) |
| Noometry Index | 43.4 | 31.9 |
| Released | 2026-04-22 | 2025-04-29 |
| Weights | Open | Open |
| Context window | 1.05M | — |
| Max output | 131K | — |
| Input $ / M tokens | $0.14 | — |
| Output $ / M tokens | $0.28 | — |
| Results tracked | 23 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
MiMo-V2.5: 43.9 (#81), Qwen3-4B: —
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| LMArena WebDev | 1438 | — |
| SciCode | 43.1% | — |
| LMArena Coding | 1469 | — |
| ALE-Bench | 513.95 | — |
Agentic & Tool Use Not comparable
MiMo-V2.5: —, Qwen3-4B: 27.6 (#100)
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 35.7% |
Reasoning MiMo-V2.5 leads
MiMo-V2.5: 28.6 (#101), Qwen3-4B: 19.2 (#268)
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| CritPt | 3.7% | — |
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | 1450 | — |
Math MiMo-V2.5 leads
MiMo-V2.5: 36.8 (#163), Qwen3-4B: 29.7 (#240)
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| MathArena Final-Answer Competitions | — | 38.5% |
| OTIS Mock AIME 2024-2025 | — | 52.2% |
| ProofBench | 16% | — |
| LMArena Math | 1436 | — |
Knowledge MiMo-V2.5 leads
MiMo-V2.5: 40.8 (#115), Qwen3-4B: 33.0 (#208)
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| GPQA Diamond | — | 52.3% |
| Vectara Hallucination Rate | — | 5.7% |
| LMArena Expert | 1460 | — |
Multimodal Not comparable
MiMo-V2.5: 39.8 (#54), Qwen3-4B: —
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| LMArena Vision | 1247 | — |
Multilingual Not comparable
MiMo-V2.5: 51.9 (#99), Qwen3-4B: —
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| LMArena Non-English | 1404 | — |
| LMArena Chinese | 1468 | — |
| LMArena French | 1447 | — |
| LMArena German | 1421 | — |
| LMArena Japanese | 1306 | — |
| LMArena Korean | 1363 | — |
| LMArena Russian | 1395 | — |
| LMArena Spanish | 1416 | — |
Instruction Following Not comparable
MiMo-V2.5: 75.5 (#60), Qwen3-4B: —
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| LMArena Instruction Following | 1434 | — |
Long Context Not comparable
MiMo-V2.5: 44.2 (#73), Qwen3-4B: —
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| LMArena Longer Query | 1445 | — |
Writing & Preference Not comparable
MiMo-V2.5: 61.6 (#86), Qwen3-4B: —
| Benchmark | MiMo-V2.5 | Qwen3-4B |
|---|---|---|
| LMArena Text | 1428 | — |
| LMArena Creative Writing | 1393 | — |
| LMArena Multi-Turn | 1445 | — |
Frequently asked questions
Is MiMo-V2.5 better than Qwen3-4B?
MiMo-V2.5 is the stronger model overall, scoring 43.4 to 31.9 on the Noometry Index.
How many benchmarks do MiMo-V2.5 and Qwen3-4B share?
0 benchmarks have published results for both models. MiMo-V2.5 has 23 scored results on Noometry and Qwen3-4B has 6.