Gemini 3.5 Flash vs GLM 5.2: Benchmark Comparison
Gemini 3.5 Flash has the higher score on 8 of 16 shared benchmarks; GLM 5.2 leads on 8.
The largest observed score gap is 18.10 pts on LiveCodeBench , where Gemini 3.5 Flash leads.
Reported ±1 standard-error ranges overlap on 6 of 16 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 | GLM 5.2 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | 2.50% ±0.83 | 7.08% ±2.00 | 4.58 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 30.77% ±3.21 | 31.25% ±3.22 | 0.48 pts | Reported ±1 SE ranges overlap |
| LegalBench | 83.60% ±0.85 | 84.07% ±0.45 | 0.47 pts | Reported ±1 SE ranges overlap |
| EMB | 63.55% ±2.75 | 61.53% ±2.88 | 2.02 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 57.86% ±0.23 | 49.70% ±0.88 | 8.16 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 74.37% ±0.85 | 73.34% ±0.86 | 1.02 pts | Reported ±1 SE ranges overlap |
| MedCode | 55.83% ±2.11 | 40.77% ±2.17 | 15.05 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 76.57% ±1.92 | 83.53% ±2.00 | 6.96 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 92.68% ±1.46 | 85.61% ±1.77 | 7.07 pts | Reported ±1 SE ranges do not overlap |
| MMLU Pro | 89.52% ±0.31 | 86.71% ±0.33 | 2.80 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 26.75% ±4.10 | 37.87% ±4.14 | 11.12 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 87.60% ±0.95 | 69.50% ±1.18 | 18.10 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.00% ±0.00 | 0.50% ±0.50 | 0.50 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 52.74% ±4.47 | 45.08% ±4.39 | 7.66 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 78.80% ±1.83 | 82.80% ±1.69 | 4.00 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 48.68% ±4.73 | 63.96% ±4.79 | 15.28 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Gemini 3.5 Flash average | GLM 5.2 average |
|---|---|---|
| Legal | 38.96% | 40.80% |
| Finance | 65.26% | 61.52% |
| Healthcare | 66.20% | 62.15% |
| Academic | 91.10% | 86.16% |
| Coding | 49.09% | 49.95% |
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 | Gemini 3.5 Flash cost | GLM 5.2 cost | Gemini 3.5 Flash latency | GLM 5.2 latency |
|---|---|---|---|---|
| Harvey's Legal Agent Benchmark | $2.12 | $2.06 | 7m58s | 20m51s |
| Legal Research Bench | $1.31 | $0.89 | 4m38s | 17m03s |
| LegalBench | N/A | N/A | 34.58s | 5.77s |
| EMB | $5.17 | $4.47 | 18m44s | 58m52s |
| Finance Agent (v2) | $2.51 | $0.71 | 5m22s | 7m53s |
| TaxEval v2 | N/A | N/A | 19.69s | 64.94s |
| MedCode | N/A | N/A | 25.29s | 95.27s |
| MedScribe | N/A | N/A | 57.08s | 2m18s |
| GPQA Diamond | N/A | N/A | 27.66s | 3m45s |
| MMLU Pro | N/A | N/A | 26.26s | 51.64s |
| Code Migration | $5.73 | $12.64 | 19m28s | 1h54m |
| LiveCodeBench | N/A | N/A | 64.37s | 8m52s |
| ProgramBench | N/A | N/A | 18m25s | 2h06m |
| SkillsBench | $1.25 | $0.81 | 7m15s | 19m52s |
| SWE-bench | $0.95 | $0.71 | 4m14s | 11m10s |
| Vibe Code Bench v1.1 | $2.54 | $8.49 | 14m53s | 1h02m |
Results available only for Gemini 3.5 Flash
- Vals Index
- MortgageTax
- Tax Agent Bench
- ProofBench v1.1
- Terminal-Bench Science
- MMMU Pro
- SAGE
- Terminal-Bench 4.0
- Public Benefits Bench v1.1
Results available only for GLM 5.2
- SRE Bench