Zh
glm-4.7-flash
Benchmark & Insights
Zhipu AI Cloudflare Workers AI
Updated Oct 4, 2026 All models
Sample size
232 runs
in window
Accuracy
65.7%
reference match · 251d
Confidence
97%
over 252 runs
Window end
Oct 4, 2026
most recent run
Input price
$0.06/MTok
prompt tokens
Output price
$0.40/MTok
completion tokens
Model insights
- 01 Errs toward alarm: it misses few "unsafe" days (4 of 25) but called 29 "MEDIUM_SAFE" and 10 "LOW" days "unsafe".
- 02 It is also off by two levels more often than most.
- 03 Very cheap ($0.06/$0.40), but too noisy for daily use.
Recent forecasts