Model comparison
DeepSeek-V3 vs MiMo-V2-Pro
MiMo-V2-Pro is the stronger model overall, scoring 43.0 to 39.5 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. DeepSeek-V3 scores higher in 0 categories and MiMo-V2-Pro in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where MiMo-V2-Pro leads 41.5 to 34.0.
- DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $0.43 / $0.87 for MiMo-V2-Pro.
- MiMo-V2-Pro accepts more context: 1.05M tokens versus 164K.
- DeepSeek-V3 has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3 | MiMo-V2-Pro | |
|---|---|---|
| Provider | DeepSeek | Xiaomi |
| Noometry Index | 39.5 | 43.0 |
| Released | 2024-12-26 | 2026-03-18 |
| Weights | Open | Proprietary |
| Context window | 164K | 1.05M |
| Max output | 164K | 131K |
| Input $ / M tokens | $0.24 | $0.43 |
| Output $ / M tokens | $0.90 | $0.87 |
| Results tracked | 60 | 23 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiMo-V2-Pro leads
DeepSeek-V3: 42.3 (#106), MiMo-V2-Pro: 43.8 (#83)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Coding | 1368 | 1476 |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1433 |
| SciCode | 35.8% | — |
| WeirdML | 36.1% | — |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| ALE-Bench | — | 785.17 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, MiMo-V2-Pro: —
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| METR Time Horizons | 49.6% | — |
Reasoning MiMo-V2-Pro leads
DeepSeek-V3: 20.5 (#236), MiMo-V2-Pro: 22.1 (#206)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Hard Prompts | 1365 | 1457 |
| SimpleBench | 27.2% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 25.8% |
| CritPt | 0% | — |
| Thematic Generalization | — | 45.9% |
| LiveBench Reasoning | 65.8% | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| BIG-Bench Hard | 87.5% | — |
| Epoch Capabilities Index | 135.94 | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math MiMo-V2-Pro leads
DeepSeek-V3: 32.1 (#219), MiMo-V2-Pro: 39.5 (#102)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Math | 1373 | 1447 |
| OTIS Mock AIME 2024-2025 | 37.8% | — |
| Omni-MATH | 40.3% | — |
| LiveBench Math | 73.5% | — |
| MATH Level 5 | 75.5% | — |
| FrontierMath (Feb 2025 set) | 1.7% | — |
Knowledge MiMo-V2-Pro leads
DeepSeek-V3: 37.5 (#155), MiMo-V2-Pro: 41.4 (#111)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Expert | 1351 | 1478 |
| GPQA Diamond | 67.6% | — |
| MMLU-Pro | 72.3% | — |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 53.8% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |
Multilingual MiMo-V2-Pro leads
DeepSeek-V3: 48.5 (#143), MiMo-V2-Pro: 52.7 (#81)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Non-English | 1358 | 1416 |
| LMArena Chinese | 1391 | 1456 |
| LMArena French | 1385 | 1469 |
| LMArena German | 1374 | 1417 |
| LMArena Japanese | 1333 | 1366 |
| LMArena Korean | 1319 | 1400 |
| LMArena Russian | 1373 | 1427 |
| LMArena Spanish | 1358 | 1457 |
Instruction Following MiMo-V2-Pro leads
DeepSeek-V3: 72.8 (#130), MiMo-V2-Pro: 76.0 (#49)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Instruction Following | 1345 | 1445 |
| LiveBench Instruction Following | 81.5% | — |
| IFEval | 83.2% | — |
Long Context MiMo-V2-Pro leads
DeepSeek-V3: 34.0 (#253), MiMo-V2-Pro: 41.5 (#138)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Longer Query | 1352 | 1455 |
| Fiction.LiveBench | 50% | — |
| CL-bench | — | 15.7% |
| CL-bench Life | — | 6.9% |
Writing & Preference MiMo-V2-Pro leads
DeepSeek-V3: 57.4 (#130), MiMo-V2-Pro: 62.8 (#70)
| Benchmark | DeepSeek-V3 | MiMo-V2-Pro |
|---|---|---|
| LMArena Text | 1375 | 1436 |
| LMArena Creative Writing | 1364 | 1415 |
| LMArena Multi-Turn | 1389 | 1456 |
| Short-Story Creative Writing | 77% | — |
| EQ-Bench Creative Writing | 1472 | — |
| WildBench | 83% | — |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than MiMo-V2-Pro?
MiMo-V2-Pro is the stronger model overall, scoring 43.0 to 39.5 on the Noometry Index.
Which is cheaper, DeepSeek-V3 or MiMo-V2-Pro?
DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; MiMo-V2-Pro lists at $0.43 and $0.87.
Is DeepSeek-V3 or MiMo-V2-Pro better for coding?
MiMo-V2-Pro scores higher on coding benchmarks: 43.8 versus 42.3 in the Noometry coding category.
Which has the bigger context window?
MiMo-V2-Pro does, with 1.05M tokens against 164K.
How many benchmarks do DeepSeek-V3 and MiMo-V2-Pro share?
17 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and MiMo-V2-Pro has 23.