Gemini 3.5 Flash Lite vs MiniMax-M3: Benchmark Comparison
Gemini 3.5 Flash Lite has the higher score on 3 of 20 shared benchmarks; MiniMax-M3 leads on 16.
The largest observed score gap is 16.37 pts on MedScribe , where MiniMax-M3 leads.
Reported ±1 standard-error ranges overlap on 8 of 20 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 | MiniMax-M3 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 0.00% ±0.00 | 4.17% ±1.65 | 4.17 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 13.94% ±2.41 | 29.81% ±3.18 | 15.87 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 84.07% ±0.44 | 85.42% ±0.44 | 1.34 pts | Reported ±1 SE ranges do not overlap |
| EMB | 43.23% ±3.10 | 47.76% ±2.95 | 4.53 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 47.44% ±0.51 | 48.27% ±0.44 | 0.83 pts | Reported ±1 SE ranges overlap |
| MortgageTax | 68.68% ±0.91 | 68.36% ±0.91 | 0.32 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 37.45% ±3.12 | 49.69% ±3.25 | 12.24 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 72.61% ±0.88 | 72.73% ±0.86 | 0.12 pts | Reported ±1 SE ranges overlap |
| MedCode | 43.49% ±1.95 | 46.29% ±2.10 | 2.80 pts | Reported ±1 SE ranges overlap |
| MedScribe | 70.89% ±2.03 | 87.25% ±1.96 | 16.37 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 83.84% ±1.85 | 92.68% ±1.44 | 8.84 pts | Reported ±1 SE ranges do not overlap |
| MMLU Pro | 85.84% ±0.34 | 84.22% ±0.36 | 1.62 pts | Reported ±1 SE ranges do not overlap |
| MMMU Pro | 83.58% ±0.89 | 81.16% ±0.94 | 2.43 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 47.30% ±3.38 | 50.57% ±3.44 | 3.27 pts | Reported ±1 SE ranges overlap |
| Code Migration | 6.15% ±1.71 | 19.93% ±3.94 | 13.78 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 79.01% ±1.10 | 82.15% ±1.05 | 3.14 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 75.00% ±1.94 | 75.00% ±1.94 | 0.00 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 37.16% ±4.63 | 47.57% ±5.44 | 10.41 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 59.11% ±6.17 | 65.00% ±6.10 | 5.89 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 51.83% ±1.30 | 64.14% ±1.25 | 12.31 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Gemini 3.5 Flash Lite average | MiniMax-M3 average |
|---|---|---|
| Legal | 32.67% | 39.80% |
| Finance | 53.88% | 57.36% |
| Healthcare | 57.19% | 66.77% |
| Academic | 84.42% | 86.02% |
| Education | 47.30% | 50.57% |
| Coding | 49.33% | 56.16% |
| Cyber | 59.11% | 65.00% |
| Social Mobility | 51.83% | 64.14% |
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 | MiniMax-M3 cost | Gemini 3.5 Flash Lite latency | MiniMax-M3 latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $0.33 | $1.46 | 3m53s | 22m31s |
| Legal Research Bench | $0.12 | $0.34 | 59.20s | 13m34s |
| LegalBench | N/A | N/A | 1.67s | 8.21s |
| EMB | $1.23 | $2.09 | 8m14s | 31m30s |
| Finance Agent (v2) | $0.39 | $0.32 | 2m23s | 8m17s |
| MortgageTax | N/A | N/A | 4.88s | 26.76s |
| Tax Agent Bench | $0.08 | $0.16 | 47.46s | 5m15s |
| TaxEval v2 | N/A | N/A | 8.81s | 94.04s |
| MedCode | N/A | N/A | 6.02s | 63.12s |
| MedScribe | N/A | N/A | 18.99s | 2m04s |
| GPQA Diamond | N/A | N/A | 13.85s | 4m39s |
| MMLU Pro | N/A | N/A | 4.82s | 41.39s |
| MMMU Pro | N/A | N/A | 7.80s | 68.35s |
| SAGE | N/A | N/A | 10.75s | 118.54s |
| Code Migration | $0.53 | $7.07 | 22m02s | 1h14m |
| LiveCodeBench | N/A | N/A | 21.08s | 6m07s |
| SWE-bench | $0.31 | $0.42 | 7m20s | 12m06s |
| Vibe Code Bench v1.1 | $0.83 | $6.45 | 8m22s | 1h25m |
| CyberBench v1.1 | $0.43 | $3.41 | 17m56s | 40m56s |
| Public Benefits Bench v1.1 | $0.10 | $0.25 | 69.26s | 11m31s |
Results available only for Gemini 3.5 Flash Lite
- ProgramBench
Results available only for MiniMax-M3
- Vals Index
- ProofBench v1.1
- SkillsBench
- Terminal-Bench 4.0
- Vibe Code Bench 1-100