GLM-4.7 (Reasoning): benchmark results
GLM-4.7 evaluated with reasoning enabled. Provider: Zhipu. Released 2025-12-22. Access: Open.
Aggregate Rating 1672 ± 18, rank #448 of 2168 rated models, from 74 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BALSAM - Sequence Tagging | 46.49 | Overall score (0-100, LLM-judged generation and multiple cho | 100 |
| AA LiveCodeBench | 89.42 | Pass@1 (%) | 99 |
| AA AIME 2025 | 95 | Accuracy (%) | 96.7 |
| BALSAM - Fill in the Blank | 82.88 | Overall score (0-100, LLM-judged generation and multiple cho | 96.3 |
| BALSAM - Information Extraction | 53.15 | Overall score (0-100, LLM-judged generation and multiple cho | 96.3 |
| Tau2-Bench Telecom | 95.91 | Success Rate (%) | 96.3 |
| AA TAU-2 Bench | 95.91 | Accuracy (%) | 96.2 |
| AA MMLU-Pro | 85.61 | Accuracy (%) | 92.8 |
| BALSAM - Creative Writing | 52.76 | Overall score (0-100, LLM-judged generation and multiple cho | 92.6 |
| BALSAM - Logic | 40.51 | Overall score (0-100, LLM-judged generation and multiple cho | 92.6 |
| BALSAM - Question Answering | 73.95 | Overall score (0-100, LLM-judged generation and multiple cho | 92.6 |
| BALSAM - Overall | 58.01 | Mean of category overall scores (0-100) | 92.3 |
Interactive version: theaggregate.ai/model?slug=glm-4-7-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.