Google released Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, available the same day through the Gemini API, Gemini Enterprise and the Gemini app. Google says 3.6 Flash costs less per token than 3.5 Flash while using fewer tokens to finish the same work, and scores higher on coding, document and knowledge-work tests. A third model tuned for finding and fixing software security flaws will be restricted to governments and trusted partners in a limited pilot.
What changed
Gemini 3.5 Flash was the workhorse option, at a higher price per token and using more output tokens for the same tasks.
What it unlocks
Running document parsing, chart and data analysis, report drafting and multi-step automated workflows at lower cost per task, with a very fast low-cost model for high-volume jobs.
- 3.6 Flash: $1.50/1M input tokens, $7.50/1M output tokens
- 17% fewer output tokens than 3.5 Flash (Artificial Analysis Index), up to 65% on DeepSWE
- 3.5 Flash-Lite: 350 output tokens per second, $0.30/1M input and $2.50/1M output
- DeepSWE 49% vs 37%; MLE Bench 63.9% vs 49.7%; OSWorld-Verified 83.0% vs 78.4%
What you need to act on it
- Gemini API via Google AI Studio or Android Studio for developers
- Gemini Enterprise Agent Platform for enterprises
- 3.5 Flash Cyber limited to governments and trusted partners through a CodeMender pilot
Sources