Google Announces "Gemini 3.6 Flash," "3.5 Flash-Lite," and "3.5 Flash Cyber"
Google has announced a new lineup of models built for real-world agentic deployment: Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber. Gemini 3.6 Flash began rolling out the same day to third-party products such as GitHub Copilot, positioning it as the new workhorse of the Flash series.
Details
- Gemini 3.6 Flash: The primary workhorse model, with improved coding, knowledge-work, and multimodal capabilities. It cuts output token usage by 17% versus 3.5 Flash, and up to 65% fewer tokens on some benchmarks like DeepSWE. Scores 83.0% on OSWorld-Verified (up from 78.4%) and 63.9% on MLE Bench (up from 49.7%). Priced at $1.50/1M input tokens and $7.50/1M output tokens — cheaper than 3.5 Flash while performing better
- Gemini 3.5 Flash-Lite: The fastest, most cost-effective option in the 3.5 class, reaching 350 output tokens per second and significantly outperforming prior Flash-Lite generations on agentic workflows. Scores 54% on Terminal-Bench 2.1 (up from 31%). Priced at $0.30/1M input tokens and $2.50/1M output tokens
- Gemini 3.5 Flash Cyber: A specialized cybersecurity model paired with “CodeMender” agents for detecting and fixing vulnerabilities. Available only through a limited-access pilot program for governments and trusted partners — not generally available
- Availability: Gemini 3.6 Flash and Gemini 3.5 Flash-Lite launched the same day in Google AI Studio, Android Studio, the Gemini Enterprise Agent Platform, and the Gemini app. Flash-Lite is also rolling out in Google Search
How to try it
- Try Gemini 3.6 Flash and Gemini 3.5 Flash-Lite now through Google AI Studio or the Gemini app
- Gemini 3.6 Flash is also becoming available through third-party products such as GitHub Copilot
- Gemini 3.5 Flash Cyber is not generally available and is limited to a pilot program for governments and trusted partners