MiMo V2.5 Pro vs GLM 5.3 Flash: Benchmark Comparison
MiMo V2.5 Pro has the higher score on 4 of 15 shared benchmarks; GLM 5.3 Flash leads on 11.
The largest observed score gap is 29.33 pts on Legal Research Bench , where GLM 5.3 Flash leads.
Reported ±1 standard-error ranges overlap on 6 of 15 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 | MiMo V2.5 Pro | GLM 5.3 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 2.08% ±0.83 | 6.67% ±1.65 | 4.58 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 15.87% ±2.54 | 45.19% ±3.46 | 29.33 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 77.13% ±0.55 | 83.93% ±0.43 | 6.80 pts | Reported ±1 SE ranges do not overlap |
| EMB | 55.23% ±3.11 | 55.93% ±3.56 | 0.70 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 41.50% ±1.19 | 57.85% ±2.06 | 16.35 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 34.79% ±2.62 | 61.96% ±3.26 | 27.17 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 73.79% ±0.86 | 75.59% ±0.84 | 1.80 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 83.73% ±2.06 | 88.94% ±1.91 | 5.21 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 22.00% ±4.16 | 21.00% ±4.09 | 1.00 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 82.58% ±2.47 | 86.36% ±2.13 | 3.79 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 84.59% ±0.45 | 86.06% ±0.34 | 1.46 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 21.56% ±4.13 | 20.52% ±4.21 | 1.04 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 81.35% ±1.07 | 80.51% ±1.05 | 0.84 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 74.00% ±1.96 | 92.00% ±1.21 | 18.00 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 34.11% ±4.53 | 30.76% ±5.28 | 3.35 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | MiMo V2.5 Pro average | GLM 5.3 Flash average |
|---|---|---|
| Legal | 31.69% | 45.26% |
| Finance | 51.33% | 62.83% |
| Healthcare | 83.73% | 88.94% |
| Math | 22.00% | 21.00% |
| Academic | 83.59% | 86.21% |
| Coding | 52.76% | 55.95% |
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 | MiMo V2.5 Pro cost | GLM 5.3 Flash cost | MiMo V2.5 Pro latency | GLM 5.3 Flash latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $0.11 | $0.57 | 9m04s | 52m17s |
| Legal Research Bench | $0.08 | $0.12 | 6m56s | 51m07s |
| LegalBench | N/A | N/A | 27.14s | 12.52s |
| EMB | $0.22 | $0.71 | 29m08s | 1h15m |
| Finance Agent (v2) | $0.21 | $0.05 | 8m17s | 14m51s |
| Tax Agent Bench | $0.04 | $0.14 | 6m17s | 21m37s |
| TaxEval v2 | N/A | N/A | 62.83s | 53.06s |
| MedScribe | N/A | N/A | 90.71s | 86.98s |
| ProofBench v1.1 | $0.09 | $0.22 | 19m37s | 1h11m |
| GPQA Diamond | N/A | N/A | 4m11s | 2m11s |
| MMLU Pro | N/A | N/A | 67.76s | 44.76s |
| Code Migration | $0.24 | $3.13 | 48m53s | 3h26m |
| LiveCodeBench | N/A | N/A | 4m31s | 5m02s |
| SWE-bench | $0.02 | $0.02 | 7m51s | 1h04m |
| Vibe Code Bench v1.1 | $0.17 | $2.35 | 40m04s | 1h24m |
Results available only for MiMo V2.5 Pro
- Vals Index
- MedCode
- Terminal-Bench 4.0
Results available only for GLM 5.3 Flash
- Terminal-Bench Science
- MMMU Pro
- IOI
- ProgramBench
- SkillsBench
- Vibe Code Bench 1-100
- CyberBench v1.1