Google launches Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber, with cheaper and more token-efficient Flash series

Gemini 3.6 Flash, the new default work model, reduces output tokens by 17% vs 3.5 Flash (up to 65% on DeepSWE), with output price dropping from $9.00 to $7.50/M tokens. It scores 49% vs 37% on DeepSWE, 63.9% vs 49.7% on MLE Bench, and 83.0% vs 78.4% on OSWorld-Verified. Computer use is available as a built-in client tool via Gemini API and Gemini Enterprise. Gemini 3.5 Flash-Lite targets low-latency high-throughput scenarios at 350 tokens/s, priced $0.30/M input and $2.50/M output, achieving 54% on Terminal-Bench 2.1, with configurable thinking levels. Gemini 3.5 Flash Cyber, fine-tuned from 3.5 Flash, specializes in discovering, verifying, and patching software vulnerabilities, using up to five parallel calls in CodeMender. On Big Sleep evaluation, it found 55 unique confirmed issues in V8 engine, surpassing 3.5 Flash's 47 and Claude Opus 4.6's 36. Due to dual-use risk, Flash Cyber access is limited to government and trusted partners. Developers using Gemini API can directly call 3.6 Flash and 3.5 Flash-Lite; the latter suits high-concurrency scenarios with lower token cost and high speed. Teams building code security agents can leverage Flash Cyber for rapid vulnerability discovery in CodeMender. Enterprise users can utilize these models on the Gemini Enterprise Agent Platform.