MiniMax-M3 vs MiMo V2.5 Pro: Benchmark Comparison
MiniMax-M3 has the higher score on 13 of 18 shared benchmarks; MiMo V2.5 Pro leads on 5.
The largest observed score gap is 14.90 pts on Tax Agent Bench , where MiniMax-M3 leads.
Reported ±1 standard-error ranges overlap on 9 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 | MiniMax-M3 | MiMo V2.5 Pro | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 36.53% ±1.17 | 33.84% ±1.17 | 2.69 pts | Reported ±1 SE ranges do not overlap |
| Harvey's Legal Agent Benchmark | 4.17% ±1.65 | 2.08% ±0.83 | 2.08 pts | Reported ±1 SE ranges overlap |
| Legal Research Bench | 29.81% ±3.18 | 15.87% ±2.54 | 13.94 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 85.42% ±0.44 | 77.13% ±0.55 | 8.29 pts | Reported ±1 SE ranges do not overlap |
| EMB | 47.76% ±2.95 | 55.23% ±3.11 | 7.47 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 48.27% ±0.44 | 41.50% ±1.19 | 6.77 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 49.69% ±3.25 | 34.79% ±2.62 | 14.90 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 72.73% ±0.86 | 73.79% ±0.86 | 1.06 pts | Reported ±1 SE ranges overlap |
| MedCode | 46.29% ±2.10 | 32.48% ±1.91 | 13.80 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 87.25% ±1.96 | 83.73% ±2.06 | 3.52 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 18.00% ±3.86 | 22.00% ±4.16 | 4.00 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 92.68% ±1.44 | 82.58% ±2.47 | 10.10 pts | Reported ±1 SE ranges do not overlap |
| MMLU Pro | 84.22% ±0.36 | 84.59% ±0.45 | 0.38 pts | Reported ±1 SE ranges overlap |
| Code Migration | 19.93% ±3.94 | 21.56% ±4.13 | 1.63 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 82.15% ±1.05 | 81.35% ±1.07 | 0.80 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 75.00% ±1.94 | 74.00% ±1.96 | 1.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 1.01% ±1.01 | 0.51% ±0.51 | 0.51 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 47.57% ±5.44 | 34.11% ±4.53 | 13.45 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | MiniMax-M3 average | MiMo V2.5 Pro average |
|---|---|---|
| Legal | 39.80% | 31.69% |
| Finance | 54.61% | 51.33% |
| Healthcare | 66.77% | 58.11% |
| Math | 18.00% | 22.00% |
| Academic | 88.45% | 83.59% |
| Coding | 45.13% | 42.31% |
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 | MiniMax-M3 cost | MiMo V2.5 Pro cost | MiniMax-M3 latency | MiMo V2.5 Pro latency |
|---|---|---|---|---|
| Vals Index | $3.08 | $0.16 | 38m55s | 34m33s |
| Harvey's Legal Agent Benchmark | $1.46 | $0.11 | 22m31s | 9m04s |
| Legal Research Bench | $0.34 | $0.08 | 13m34s | 6m56s |
| LegalBench | N/A | N/A | 8.21s | 27.14s |
| EMB | $2.09 | $0.22 | 31m30s | 29m08s |
| Finance Agent (v2) | $0.32 | $0.21 | 8m17s | 8m17s |
| Tax Agent Bench | $0.16 | $0.04 | 5m15s | 6m17s |
| TaxEval v2 | N/A | N/A | 94.04s | 62.83s |
| MedCode | N/A | N/A | 63.12s | 38.45s |
| MedScribe | N/A | N/A | 2m04s | 90.71s |
| ProofBench v1.1 | $0.42 | $0.09 | 11m59s | 19m37s |
| GPQA Diamond | N/A | N/A | 4m39s | 4m11s |
| MMLU Pro | N/A | N/A | 41.39s | 67.76s |
| Code Migration | $7.07 | $0.24 | 1h14m | 48m53s |
| LiveCodeBench | N/A | N/A | 6m07s | 4m31s |
| SWE-bench | $0.42 | $0.02 | 12m06s | 7m51s |
| Terminal-Bench 4.0 | $7.23 | $0.23 | 1h16m | 2h07m |
| Vibe Code Bench v1.1 | $6.45 | $0.17 | 1h25m | 40m04s |
Results available only for MiniMax-M3
- MortgageTax
- MMMU Pro
- SAGE
- SkillsBench
- Vibe Code Bench 1-100
- CyberBench v1.1
- Public Benefits Bench v1.1
Results available only for MiMo V2.5 Pro
None.