MiMo V2 Flash Base is the stronger all-rounder.
Scores updated · 41 tests both models report · How we compare
Where each one wins
Tests won in each of the two areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
Coding and reasoning rest on a single test each.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where MiMo V2 Flash Base pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+23.3points ahead
- Recent programming contest problemsLiveCodeBench v6+5.9points ahead
Where DeepSeek-V3.2-Exp-Base pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 41 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningMiMo V2 Flash Base
- GPQA DiamondMiMo V2 Flash Base by 23.331.855.1+23.3
Other results39 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- HumanEvalMiMo V2 Flash Base by 15.561.977.4+15.5
- NIAH-Multi (64K)MiMo V2 Flash Base by 1485.999.9+14
- MultiPL-E HumanEvalMiMo V2 Flash Base by 13.845.759.5+13.8
- NIAH-Multi (32K)MiMo V2 Flash Base by 13.785.699.3+13.7
- GSM-Infinite Hard (16K)DeepSeek-V3.2-Exp-Base by 12.750.437.7+12.7
- GSM-Infinite Hard (32K)DeepSeek-V3.2-Exp-Base by 11.545.233.7+11.5
- MATHMiMo V2 Flash Base by 10.960.171+10.9
- aime_2024_2025_combinedMiMo V2 Flash Base by 10.524.835.3+10.5
- MMLU-ProMiMo V2 Flash Base by 9.963.373.2+9.9
- GSM8KMiMo V2 Flash Base by 7.984.492.3+7.9
- BigCodeBenchMiMo V2 Flash Base by 7.262.970.1+7.2
- Chinese SimpleQA (C-SimpleQA)DeepSeek-V3.2-Exp-Base by 6.56861.5+6.5
- SimpleQADeepSeek-V3.2-Exp-Base by 6.42720.6+6.4
- GPQA (unspecified)MiMo V2 Flash Base by 6.237.343.5+6.2
- MultiPL-E MBPPMiMo V2 Flash Base by 6.150.656.7+6.1
- IncludeDeepSeek-V3.2-Exp-Base by 5.877.271.4+5.8
- GlobalMMLUDeepSeek-V3.2-Exp-Base by 5.48276.6+5.4
- MBPPMiMo V2 Flash Base by 5.475.681+5.4
- NIAH-Multi (128K)MiMo V2 Flash Base by 4.394.398.6+4.3
- CRUXEval-O (output prediction)MiMo V2 Flash Base by 4.274.979.1+4.2
- CRUXEval-I (input prediction)MiMo V2 Flash Base by 3.663.967.5+3.6
- TriviaQADeepSeek-V3.2-Exp-Base by 3.683.980.3+3.6
- GSM-Infinite Hard (Tools-allowed)MiMo V2 Flash Base by 3.325.729+3.3
- GSM-Infinite Hard (128K)MiMo V2 Flash Base by 3.325.729+3.3
- C-EvalDeepSeek-V3.2-Exp-Base by 3.19187.9+3.1
- HumanEval+MiMo V2 Flash Base by 367.770.7+3
- SuperGPQADeepSeek-V3.2-Exp-Base by 2.543.641.1+2.5
- NIAH-Multi (Tools-allowed)MiMo V2 Flash Base by 2.494.396.7+2.4
- DROPDeepSeek-V3.2-Exp-Base by 1.986.684.7+1.9
- MBPP+MiMo V2 Flash Base by 1.669.871.4+1.6
- MBPP+ (EvalPlus-augmented)MiMo V2 Flash Base by 1.669.871.4+1.6
- CMMLUDeepSeek-V3.2-Exp-Base by 1.588.987.4+1.5
- GSM-Infinite Hard (64K)DeepSeek-V3.2-Exp-Base by 1.132.631.5+1.1
- MMLUDeepSeek-V3.2-Exp-Base by 1.187.886.7+1.1
- HellaSwagtie89.488.5tie
- ARC-Challengetie95.295.9tie
- WinoGrandetie83.483.8tie
- BBHtie88.788.5tie
- MMLU-Reduxtie90.490.6tie
Questions people ask
Which is better, DeepSeek-V3.2-Exp-Base or MiMo V2 Flash Base?
MiMo V2 Flash Base wins all two areas where both have results: coding and reasoning. DeepSeek-V3.2-Exp-Base wins none.
Which is better for coding?
MiMo V2 Flash Base. It wins the one coding test both models report.
How do you compare the two?
We use the 41 benchmark tests both models have published scores on. The verdict counts the 2 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 39 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.