GLM-4.6V (Reasoning) — benchmark results
GLM-4.6V evaluated with reasoning enabled. Provider: Zhipu. Released 2025-12-09. Access: Open.
Unified ELO 1585 ± 28, rank #516 of 1776 rated models, from 56 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA AIME 2025 | 85.33 | Accuracy (%) | 83.3 |
| UGI - Writing | 39.66 | Writing Score | 72.8 |
| UGI - Natural Intelligence | 28.9 | NatInt Score | 71.9 |
| AA MMLU-Pro | 79.89 | Accuracy (%) | 66.6 |
| AA Omniscience | -26.48 | Score | 66.1 |
| AA Omniscience - Software Engineering (SWE) - Rust | 54 | Accuracy (%) | 59.5 |
| AA Humanity's Last Exam | 8.85 | Accuracy (%) | 58.1 |
| AA GPQA Diamond | 71.92 | Accuracy (%) | 56.1 |
| AA Global-MMLU-Lite - English | 90.25 | Accuracy (%) | 53.8 |
| Artificial Analysis Intelligence Index | 16.75 | Intelligence Index | 52.8 |
| AA Global-MMLU-Lite - Korean | 83.75 | Accuracy (%) | 52.2 |
| AA Omniscience - Software Engineering (SWE) - PHP | 24 | Accuracy (%) | 51.5 |
Interactive version: theaggregate.ai/model?slug=glm-4-6v-reasoning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.