Gemini 3.5 Flash Lite vs MiMo V2.5: Benchmark Comparison
Gemini 3.5 Flash Lite has the higher score on 12 of 18 shared benchmarks; MiMo V2.5 leads on 6.
The largest observed score gap is 11.86 pts on EMB , where MiMo V2.5 leads.
Reported ±1 standard-error ranges overlap on 5 of 18 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 | Gemini 3.5 Flash Lite | MiMo V2.5 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 0.00% ±0.00 | 1.67% ±0.00 | 1.67 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 13.94% ±2.41 | 9.13% ±2.00 | 4.81 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 84.07% ±0.44 | 78.89% ±0.51 | 5.19 pts | Reported ±1 SE ranges do not overlap |
| EMB | 43.23% ±3.10 | 55.09% ±3.13 | 11.86 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 47.44% ±0.51 | 36.73% ±0.11 | 10.71 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 68.68% ±0.91 | 59.26% ±0.98 | 9.42 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 37.45% ±3.12 | 29.16% ±2.71 | 8.29 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 72.61% ±0.88 | 71.83% ±0.88 | 0.78 pts | Reported ±1 SE ranges overlap |
| MedCode | 43.49% ±1.95 | 31.89% ±2.02 | 11.59 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 70.89% ±2.03 | 72.15% ±1.85 | 1.27 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 83.84% ±1.85 | 81.57% ±2.05 | 2.27 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 85.84% ±0.34 | 82.93% ±0.37 | 2.91 pts | Reported ±1 SE ranges do not overlap |
| MMMU Pro | 83.58% ±0.89 | 80.00% ±0.96 | 3.58 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 47.30% ±3.38 | 43.27% ±3.39 | 4.04 pts | Reported ±1 SE ranges overlap |
| Code Migration | 6.15% ±1.71 | 14.23% ±3.69 | 8.08 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 79.01% ±1.10 | 81.51% ±1.07 | 2.50 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 75.00% ±1.94 | 71.00% ±2.03 | 4.00 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 37.16% ±4.63 | 42.17% ±4.57 | 5.01 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Gemini 3.5 Flash Lite average | MiMo V2.5 average |
|---|---|---|
| Legal | 32.67% | 29.90% |
| Finance | 53.88% | 50.41% |
| Healthcare | 57.19% | 52.02% |
| Academic | 84.42% | 81.50% |
| Education | 47.30% | 43.27% |
| Coding | 49.33% | 52.23% |
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 | Gemini 3.5 Flash Lite cost | MiMo V2.5 cost | Gemini 3.5 Flash Lite latency | MiMo V2.5 latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $0.33 | $0.04 | 3m53s | 6m59s |
| Legal Research Bench | $0.12 | $0.03 | 59.20s | 6m31s |
| LegalBench | N/A | N/A | 1.67s | 11.18s |
| EMB | $1.23 | $0.08 | 8m14s | 30m38s |
| Finance Agent (v2) | $0.39 | $0.09 | 2m23s | 5m39s |
| MortgageTax | N/A | N/A | 4.88s | 32.94s |
| Tax Agent Bench | $0.08 | $0.02 | 47.46s | 8m21s |
| TaxEval v2 | N/A | N/A | 8.81s | 26.30s |
| MedCode | N/A | N/A | 6.02s | 17.70s |
| MedScribe | N/A | N/A | 18.99s | 20.26s |
| GPQA Diamond | N/A | N/A | 13.85s | 89.39s |
| MMLU Pro | N/A | N/A | 4.82s | 30.25s |
| MMMU Pro | N/A | N/A | 7.80s | 42.47s |
| SAGE | N/A | N/A | 10.75s | 52.66s |
| Code Migration | $0.53 | $0.10 | 22m02s | 37m55s |
| LiveCodeBench | N/A | N/A | 21.08s | 108.61s |
| SWE-bench | $0.31 | $0.01 | 7m20s | 4m04s |
| Vibe Code Bench v1.1 | $0.83 | $0.07 | 8m22s | 26m20s |
Results available only for Gemini 3.5 Flash Lite
- ProgramBench
- CyberBench v1.1
- Public Benefits Bench v1.1
Results available only for MiMo V2.5
- ProofBench v1.1