Up to four models
GPT-5.5 vs GLM-5.1
On the index (GPT-4 = 100), GPT-5.5 scores 172 and GLM-5.1 151. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
GPT-5.5
172156–188
WeightsClosed
ReleasedApr 2026
Price in / out$5.0 / $30
Context1.05M
Core tests taken6 of 9
151134–167
WeightsOpen
ReleasedApr 2026
Price in / out$1.4 / $4.4
Context205K
Core tests taken5 of 9
GPT-5.5 leads GLM-5.1 by 21 points, ahead on 5 of 5 shared core tests, at 5.2× the price.
Test by testthird-party results
GPT-5.5
GLM-5.1
Every result as a table
| Test | GPT-5.5 | GLM-5.1 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 94.0% | 89.9% |
| GPQA Diamond (Vals AI) | 93.2% | 84.5% |
| MMLU Pro | 88.1% | 86.9% |
| SimpleBench | 69.0% | 55.1% |
| SimpleQA Verified | 63.0% | 34.0% |
| Maths | ||
| FrontierMath (tiers 1-3) | 85.3% | 36.8% |
| AIME-style maths (OTIS mock) | 100.0% | 93.3% |
| Chess puzzles | 63.0% | 20.0% |
| FrontierMath (2025 set) | 51.7% | 33.5% |
| FrontierMath tier 4 (2025 set) | 35.4% | 12.5% |
| Coding | ||
| SWE-bench Verified | 80.6% | 74.2% |
| Code Migration | 45.2% | 25.8% |
| LiveCodeBench | 85.3% | 81.4% |
| ProgramBench | 0.5% | 0.0% |
| SWE-bench | 82.6% | 76.4% |
| Vibe Code Bench v1.1 | 69.8% | 31.5% |
| WeirdML | 84.9% | 57.1% |
| Agents | ||
| APEX-Agents | 55.1% | 40.9% |
| Professional work | ||
| Finance Agent (v2) | 51.8% | 44.8% |
| CorpFin v2 | 68.4% | 64.5% |
| LegalBench | 86.5% | 84.4% |
| Legal Research Bench | 40.4% | 27.9% |
| MedCode | 49.1% | 41.6% |
| MedScribe | 86.9% | 72.3% |
| Public Benefits Bench v1.1 | 60.9% | 61.8% |
| TaxEval v2 | 75.0% | 71.2% |
Tests at least two of these models have taken. Bold rows are the core tests.