Nv
nemotron-3-120b-a12b Benchmark & Insights
Nvidia Cloudflare Workers AI
Updated Jul 31, 2026 All models
Sample size
169 runs
in window
Accuracy
76.0%
consensus match · 204d
Confidence
90%
over 205 runs
Window end
Jul 31, 2026
most recent run
Input price
$0.50/MTok
prompt tokens
Output price
$1.50/MTok
completion tokens
Model insights
- 01 Two failure modes at once — it inflates "low" days to "medium", and separately throws "unsafe" flags on "safe" days about a dozen times.
- 02 Its risk-level score is much better than its safe score, so the flag is the part you cannot trust.
Recent forecasts