GLM-4.6V (Reasoning): benchmark results

GLM-4.6V evaluated with reasoning enabled. Provider: Zhipu. Released 2025-12-09. Access: Open.

Unified ELO 1512 ± 1, rank #793 of 1761 rated models, from 40 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
UGI - Writing39.66Writing Score72.1
UGI - Natural Intelligence28.9NatInt Score71.1
AA Omniscience-26.95Score58.7
AA Global-MMLU-Lite - English90.25Accuracy (%)53.2
AA Humanity's Last Exam9.64Accuracy (%)52.7
AA Terminal-Bench Hard14.39Accuracy (%)51.3
AA GPQA Diamond71.92Accuracy (%)51.2
AA Global-MMLU-Lite - Korean83.75Accuracy (%)50.5
Artificial Analysis Intelligence Index10.73Intelligence Index49
AA Global-MMLU-Lite - Swahili72.25Accuracy (%)48.2
AA Global-MMLU-Lite - Yoruba52.08Accuracy (%)47.7
AA Global-MMLU-Lite - French86.25Accuracy (%)46.8

Interactive version: theaggregate.ai/model?slug=glm-4-6v-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-05.