Model comparison
DeepSeek-R1 vs MiMo-V2.5-Pro
MiMo-V2.5-Pro is the stronger model overall, scoring 45.2 to 42.3 on the Noometry Index.
Last verified . 21 shared benchmarks.
DeepSeekSummary
- They share 21 benchmarks with published results for both. DeepSeek-R1 scores higher in 2 categories and MiMo-V2.5-Pro in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where MiMo-V2.5-Pro leads 26.8 to 18.6.
- The biggest single-benchmark swing is SciCode: 35.7% for DeepSeek-R1 and 50.2% for MiMo-V2.5-Pro.
- MiMo-V2.5-Pro is cheaper at $0.43 / $0.87 per million input/output tokens, against $0.50 / $2.15 for DeepSeek-R1.
- MiMo-V2.5-Pro accepts more context: 1.05M tokens versus 164K.
- MiMo-V2.5-Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-R1 | MiMo-V2.5-Pro | |
|---|---|---|
| Provider | DeepSeek | Xiaomi |
| Noometry Index | 42.3 | 45.2 |
| Released | 2025-01-20 | 2026-04-22 |
| Weights | Proprietary | Open |
| Context window | 164K | 1.05M |
| Max output | 64K | 131K |
| Input $ / M tokens | $0.50 | $0.43 |
| Output $ / M tokens | $2.15 | $0.87 |
| Results tracked | 52 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiMo-V2.5-Pro leads
DeepSeek-R1: 46.3 (#68), MiMo-V2.5-Pro: 47.4 (#60)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| SciCode | 35.7% | 50.2% |
| LMArena Coding | 1427 | 1503 |
| ALE-Bench | 804.12 | 899.8 |
| Aider Polyglot | 71.4% | — |
| LMArena WebDev | — | 1479 |
| WeirdML | 41.6% | — |
| LiveBench Coding | 66.7% | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use Not comparable
DeepSeek-R1: 30.7 (#75), MiMo-V2.5-Pro: —
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| METR Time Horizons | 53.8% | — |
Reasoning MiMo-V2.5-Pro leads
DeepSeek-R1: 18.6 (#278), MiMo-V2.5-Pro: 26.8 (#130)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| CritPt | 1.1% | 4% |
| LMArena Hard Prompts | 1416 | 1488 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 40.8% | — |
| Kagi LLM Benchmark | 69.4% | — |
| NYT Connections (extended) | — | 34.4% |
| ARC-AGI-1 | 21.2% | — |
| LiveBench Reasoning | 83.2% | — |
| DTBench | — | 84.5% |
| LiveBench Data Analysis | 69.8% | — |
| LMCA | — | 29.5% |
| Epoch Capabilities Index | 141.29 | — |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |
Math DeepSeek-R1 leads
DeepSeek-R1: 43.8 (#79), MiMo-V2.5-Pro: 40.0 (#96)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Math | 1400 | 1481 |
| OTIS Mock AIME 2024-2025 | 66.4% | — |
| ProofBench | — | 22% |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| MATH Level 5 | 96.6% | — |
Knowledge DeepSeek-R1 leads
DeepSeek-R1: 44.5 (#87), MiMo-V2.5-Pro: 42.2 (#98)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Expert | 1394 | 1503 |
| GPQA Diamond | 76.3% | — |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
Multilingual MiMo-V2.5-Pro leads
DeepSeek-R1: 52.4 (#85), MiMo-V2.5-Pro: 55.1 (#34)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Non-English | 1412 | 1449 |
| LMArena Chinese | 1442 | 1507 |
| LMArena French | 1417 | 1488 |
| LMArena German | 1404 | 1458 |
| LMArena Japanese | 1391 | 1412 |
| LMArena Korean | 1360 | 1437 |
| LMArena Russian | 1423 | 1450 |
| LMArena Spanish | 1411 | 1471 |
Instruction Following MiMo-V2.5-Pro leads
DeepSeek-R1: 72.0 (#143), MiMo-V2.5-Pro: 77.5 (#21)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Instruction Following | 1382 | 1477 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |
Long Context Too close to call
DeepSeek-R1: 45.4 (#36), MiMo-V2.5-Pro: 45.4 (#37)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Longer Query | 1391 | 1483 |
| Fiction.LiveBench | 75% | — |
Writing & Preference MiMo-V2.5-Pro leads
DeepSeek-R1: 61.4 (#88), MiMo-V2.5-Pro: 65.3 (#49)
| Benchmark | DeepSeek-R1 | MiMo-V2.5-Pro |
|---|---|---|
| LMArena Text | 1428 | 1465 |
| LMArena Creative Writing | 1405 | 1440 |
| EQ-Bench Creative Writing | 1500 | 1493 |
| LMArena Multi-Turn | 1405 | 1477 |
| Short-Story Creative Writing | 83% | — |
| WildBench | 82.8% | — |
| EQ-Bench 4 | — | 1208 |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than MiMo-V2.5-Pro?
MiMo-V2.5-Pro is the stronger model overall, scoring 45.2 to 42.3 on the Noometry Index.
Which is cheaper, DeepSeek-R1 or MiMo-V2.5-Pro?
MiMo-V2.5-Pro is cheaper. It lists at $0.43 per million input tokens and $0.87 per million output tokens; DeepSeek-R1 lists at $0.50 and $2.15.
Is DeepSeek-R1 or MiMo-V2.5-Pro better for coding?
MiMo-V2.5-Pro scores higher on coding benchmarks: 47.4 versus 46.3 in the Noometry coding category.
Which has the bigger context window?
MiMo-V2.5-Pro does, with 1.05M tokens against 164K.
How many benchmarks do DeepSeek-R1 and MiMo-V2.5-Pro share?
21 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and MiMo-V2.5-Pro has 27.