Google DeepMindLaunchAI ModelsJuly 21, 2026

Google launches Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

Read original source →deepmind.google

Google DeepMind released three new Flash-tier models: Gemini 3.6 Flash cuts output token usage by 17% (up to 65% on some coding benchmarks), 3.5 Flash-Lite hits 350 output tokens/sec, and 3.5 Flash Cyber targets cybersecurity. Gemini 3.5 Pro remains in testing, with Gemini 4 pre-training underway.

2 sources

More stories today

Open the live feed