TheTechno Optimist.

StorySoftware

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

Google DeepMindGemini

Google released three smaller Gemini models: 3.6 Flash, a stronger mid-tier model that also writes shorter answers; 3.5 Flash-Lite, its fastest and cheapest; and 3.5 Flash Cyber, tuned to find and patch security flaws. Google also said Gemini 3.5 Pro is still testing with partners and that pre-training of Gemini 4 has begun.

  • 3.6 Flash: $1.50/$7.50 per million input/output tokens and 17% fewer output tokens than 3.5 Flash, which lowers real cost per task
  • 3.6 Flash vs 3.5 Flash: DeepSWE 49% vs 37%, MLE Bench 63.9% vs 49.7%, OSWorld-Verified 83.0% vs 78.4%
  • 3.5 Flash-Lite: $0.30/$2.50, about 350 output tokens per second, rolling into Google Search
  • 3.5 Flash Cyber runs inside the CodeMender agent and is limited to a pilot with governments and trusted partners
Read the original · Blog AI

More on Gemini

Primer