Up to four models
GLM-5 vs GLM-4.7
On the index (GPT-4 = 100), GLM-5 scores 142 and GLM-4.7 140. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
142126–157
WeightsOpen
ReleasedFeb 2026
Price in / out$1.0 / $3.2
Context205K
Core tests taken3 of 9
140123–157
WeightsOpen
ReleasedDec 2025
Price in / out$0.40 / $1.75
Context205K
Core tests taken1 of 9
GLM-5 and GLM-4.7 are too close to call (2 points apart, within the uncertainty), ahead on 1 of 1 shared core tests, at 2.1× the price.
Test by testthird-party results
Every result as a table
| Test | GLM-5 | GLM-4.7 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 87.8% | 83.3% |
| GPQA Diamond (Vals AI) | 83.3% | 80.0% |
| MMLU Pro | 86.0% | 82.7% |
| CL-bench | 18.7% | 15.9% |
| SimpleBench | 53.2% | 47.7% |
| Maths | ||
| AIME-style maths (OTIS mock) | 80.0% | 83.3% |
| FrontierMath (2025 set) | 16.4% | 2.4% |
| FrontierMath tier 4 (2025 set) | 2.1% | 0.0% |
| Coding | ||
| Terminal-Bench 2.0 (Terminus 2 runs) | 52.4% | 33.4% |
| LiveCodeBench | 81.9% | 82.2% |
| SWE-bench | 71.4% | 69.4% |
| Professional work | ||
| CorpFin v2 | 62.9% | 46.4% |
| LegalBench | 84.1% | 83.4% |
| TaxEval v2 | 70.0% | 68.8% |
Tests at least two of these models have taken. Bold rows are the core tests.