Go
gemini-3.6-flash
Benchmark & Insights
Google Gemini API
Updated Oct 4, 2026 All models
Sample size
232 runs
in window
Accuracy
71.5%
reference match · 281d
Confidence
90%
over 281 runs
Window end
Oct 4, 2026
most recent run
Input price
$1.50/MTok
prompt tokens
Output price
$7.50/MTok
completion tokens
Model insights
- 01 Behaves almost the same as 3.8: zero false alarms, 46 "MEDIUM_SAFE" days called "low", and 13 missed "unsafe" days.
- 02 Dropped to 61% in August, a month with many "MEDIUM_SAFE" days.
- 03 Worse than the older and similarly priced gemini-3.5-flash.
Notes
20% cheaper than the previous version in output tokens. Gemini 3.6 Flash Model Card
Recent forecasts