GPT-5.6 Terra vs GLM 5.3: Benchmark Comparison
GPT-5.6 Terra has the higher score on 10 of 21 shared benchmarks; GLM 5.3 leads on 9.
The largest observed score gap is 25.00 pts on ProofBench v1.1 , where GPT-5.6 Terra leads.
Reported ±1 standard-error ranges overlap on 11 of 21 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 | GPT-5.6 Terra | GLM 5.3 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 41.35% ±3.42 | 49.04% ±3.48 | 7.69 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 85.11% ±0.45 | 84.84% ±0.40 | 0.27 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 30.04% ±3.31 | 39.70% ±3.52 | 9.66 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 76.17% ±0.84 | 72.36% ±0.88 | 3.80 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 43.41% ±2.17 | 42.86% ±2.11 | 0.55 pts | Reported ±1 SE ranges overlap |
| MedScribe | 82.87% ±1.95 | 88.81% ±2.00 | 5.94 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 74.00% ±4.41 | 49.00% ±5.02 | 25.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 10.00% ±3.61 | 5.71% ±2.79 | 4.29 pts | Reported ±1 SE ranges overlap |
| GPQA Diamond | 90.91% ±1.91 | 88.13% ±1.85 | 2.78 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 86.66% ±0.33 | 86.77% ±0.34 | 0.11 pts | Reported ±1 SE ranges overlap |
| IOI | 87.61% ±6.53 | 68.44% ±7.48 | 19.17 pts | Reported ±1 SE ranges do not overlap |
| LiveCodeBench | 85.93% ±1.02 | 80.53% ±1.06 | 5.40 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 1.50% ±0.86 | 1.00 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 58.90% ±4.47 | 47.51% ±4.52 | 11.39 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 95.40% ±0.94 | 95.40% ±0.94 | 0.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 22.73% ±2.31 | 38.89% ±1.82 | 16.16 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench 1-100 | 14.82% ±3.12 | 19.99% ±3.81 | 5.17 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 74.59% ±4.20 | 78.13% ±3.94 | 3.53 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 72.08% ±5.41 | 72.08% ±5.41 | 0.00 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 9.54% ±1.82 | 9.16% ±1.79 | 0.38 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 62.38% ±1.26 | 68.54% ±1.21 | 6.16 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | GPT-5.6 Terra average | GLM 5.3 average |
|---|---|---|
| Legal | 63.23% | 66.94% |
| Finance | 53.10% | 56.03% |
| Healthcare | 63.14% | 65.84% |
| Math | 74.00% | 49.00% |
| Science | 10.00% | 5.71% |
| Academic | 88.78% | 87.45% |
| Coding | 55.06% | 53.80% |
| Cyber | 40.81% | 40.62% |
| Social Mobility | 62.38% | 68.54% |
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 | GPT-5.6 Terra cost | GLM 5.3 cost | GPT-5.6 Terra latency | GLM 5.3 latency |
|---|---|---|---|---|
| Legal Research Bench | $7.91 | $2.24 | 1h10m | 51m50s |
| LegalBench | N/A | N/A | 2.78s | 22.48s |
| Tax Agent Bench | $4.83 | $1.80 | 1h01m | 50m53s |
| TaxEval v2 | N/A | N/A | 37.28s | 98.32s |
| MedCode | N/A | N/A | 18.41s | 2m49s |
| MedScribe | N/A | N/A | 35.53s | 2m02s |
| ProofBench v1.1 | $1.27 | $2.08 | 12m57s | 42m23s |
| Terminal-Bench-Science | $5.19 | $15.63 | 2h04m | 2h45m |
| GPQA Diamond | N/A | N/A | 36.76s | 3m31s |
| MMLU Pro | N/A | N/A | 8.17s | 61.78s |
| IOI | $8.66 | $7.67 | 1h23m | 1h46m |
| LiveCodeBench | N/A | N/A | 44.33s | 4m07s |
| ProgramBench | $5.38 | $21.96 | 33m36s | 4h16m |
| SkillsBench | $1.75 | $0.71 | 6m55s | 13m07s |
| SWE-bench | $0.40 | $0.34 | 3m00s | 14m01s |
| Terminal-Bench 4.0 | $5.60 | $9.37 | 31m29s | 1h23m |
| Vibe Code Bench 1-100 | $33.33 | $8.24 | 1h49m | 1h26m |
| Vibe Code Bench v1.1 | $7.89 | $12.45 | 19m12s | 1h04m |
| CyberBench v1.1 | $3.31 | $2.69 | 17m04s | 30m18s |
| SRE Bench | $30.86 | $27.47 | 3h17m | 2h48m |
| Public Benefits Bench v1.1 | $1.20 | $0.95 | 19m31s | 35m51s |
Results available only for GPT-5.6 Terra
- MortgageTax
- MMMU Pro
- SAGE
Results available only for GLM 5.3
- Vals Finance Index
- Vals Index
- Vals Legal Index
- Vals RSI Index
- HLAB-Vals-1.1
- EMB v1.1
- Finance Agent v2.1
- MysteryMechanism
- Code Migration v1.1