DeepSeek V4.1 Flash vs MiMo V2.6 Flash: Benchmark Comparison
DeepSeek V4.1 Flash has the higher score on 7 of 21 shared benchmarks; MiMo V2.6 Flash leads on 13.
The largest observed score gap is 19.38 pts on Code Migration v1.1 , where DeepSeek V4.1 Flash leads.
Reported ±1 standard-error ranges overlap on 15 of 21 shared benchmarks with comparable uncertainty data. This is not a pairwise statistical significance test.
Shared benchmark results
Scores and reported standard errors come from the latest visible Vals benchmark version available for both models.
| Benchmark | DeepSeek V4.1 Flash | MiMo V2.6 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Finance Index | 57.08% ±1.45 | 57.15% ±1.39 | 0.07 pts | Reported ±1 SE ranges overlap |
| Vals Science Index | 29.65% ±0.99 | 31.72% ±1.27 | 2.07 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 41.35% ±3.42 | 37.98% ±3.37 | 3.36 pts | Reported ±1 SE ranges overlap |
| EMB v1.1 | 67.76% ±2.56 | 68.88% ±2.27 | 1.12 pts | Reported ±1 SE ranges overlap |
| Finance Agent v2.1 | 65.43% ±2.02 | 63.32% ±2.08 | 2.11 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 26.63% ±3.14 | 28.51% ±3.14 | 1.88 pts | Reported ±1 SE ranges overlap |
| MedCode | 41.17% ±2.04 | 41.06% ±2.03 | 0.12 pts | Reported ±1 SE ranges overlap |
| MedScribe | 85.50% ±1.92 | 85.28% ±1.99 | 0.22 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 54.00% ±5.01 | 63.00% ±4.85 | 9.00 pts | Reported ±1 SE ranges overlap |
| BioMysteryBench | 67.78% ±1.11 | 69.26% ±0.98 | 1.48 pts | Reported ±1 SE ranges overlap |
| MysteryMechanism | 21.17% ±2.75 | 21.62% ±2.77 | 0.45 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench-Science | 0.00% ±0.00 | 4.29% ±2.44 | 4.29 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 47.88% ±3.44 | 43.53% ±3.43 | 4.35 pts | Reported ±1 SE ranges overlap |
| Code Migration v1.1 | 58.94% ±2.66 | 39.55% ±3.54 | 19.38 pts | Reported ±1 SE ranges do not overlap |
| IOI | 40.28% ±2.56 | 47.67% ±4.79 | 7.39 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 0.50% ±0.50 | 0.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 19.70% ±1.75 | 24.24% ±1.51 | 4.55 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 84.74% ±2.89 | 78.96% ±4.04 | 5.78 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 73.69% ±5.48 | 75.36% ±5.42 | 1.67 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 0.76% ±0.54 | 4.20% ±1.24 | 3.44 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 64.28% ±1.25 | 67.59% ±1.22 | 3.31 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | DeepSeek V4.1 Flash average | MiMo V2.6 Flash average |
|---|---|---|
| Legal | 41.35% | 37.98% |
| Finance | 53.27% | 53.57% |
| Healthcare | 63.34% | 63.17% |
| Math | 54.00% | 63.00% |
| Science | 29.65% | 31.72% |
| Education | 47.88% | 43.53% |
| Coding | 40.83% | 38.18% |
| Cyber | 37.23% | 39.78% |
| Social Mobility | 64.28% | 67.59% |
Cost and latency
Cost per test appears only where the benchmark reports it for both models. Latency is the measured completion time for that benchmark.
| Benchmark | DeepSeek V4.1 Flash cost | MiMo V2.6 Flash cost | DeepSeek V4.1 Flash latency | MiMo V2.6 Flash latency |
|---|---|---|---|---|
| Vals Finance Index | $0.25 | $0.07 | 11m37s | 19m26s |
| Vals Science Index | $0.29 | $0.12 | 1h07m | 1h41m |
| Legal Research Bench | $0.25 | $0.07 | 11m42s | 15m20s |
| EMB v1.1 | $0.23 | $0.11 | 13m00s | 25m22s |
| Finance Agent v2.1 | $0.37 | $0.07 | 16m00s | 15m36s |
| Tax Agent Bench | $0.13 | $0.05 | 5m49s | 17m19s |
| MedCode | N/A | N/A | 35.13s | 93.37s |
| MedScribe | N/A | N/A | 53.83s | 74.48s |
| ProofBench v1.1 | $0.13 | $0.12 | 9m07s | 1h03m |
| BioMysteryBench | $0.23 | $0.05 | 14m50s | 28m28s |
| MysteryMechanism | $0.16 | $0.06 | 8m04s | 35m08s |
| Terminal-Bench-Science | $0.47 | $0.24 | 2h57m | 3h59m |
| SAGE | N/A | N/A | 35.78s | 2m16s |
| Code Migration v1.1 | $1.09 | $0.53 | 1h54m | 4h10m |
| IOI | $0.24 | $0.22 | 18m30s | 1h49m |
| ProgramBench | $0.90 | $0.47 | 1h02m | 3h22m |
| Terminal-Bench 4.0 | $0.50 | $0.22 | 48m22s | 2h25m |
| Vibe Code Bench v1.1 | $0.41 | $0.56 | 15m28s | 50m13s |
| CyberBench v1.1 | $0.07 | $0.05 | 13m50s | 28m22s |
| SRE Bench | $0.55 | $0.40 | 1h01m | 3h22m |
| Public Benefits Bench v1.1 | $0.07 | $0.03 | 9m59s | 20m56s |
Results available only for DeepSeek V4.1 Flash
- Vals Inference Provider Index
- LegalBench
- SkillsBench
- SWE-bench Multimodal
- Vibe Code Bench 1-100
Results available only for MiMo V2.6 Flash
- Vals Index
- Vals Legal Index
- HLAB-Vals-1.1