Particle.news

Google Ships Three New Gemini Flash Models While Gemini 3.5 Pro Remains In Testing

The releases show Google is prioritizing token efficiency and lower operating costs for production AI agents.

Overview

  • On Tuesday, July 21, 2026, Google released Gemini 3.6 Flash, 3.5 Flash‑Lite, and 3.5 Flash Cyber with immediate availability for most developer and consumer channels.
  • Gemini 3.6 Flash is billed as the new workhorse and uses about 17% fewer output tokens than 3.5 Flash while lowering API output pricing to $7.50 per million tokens and showing improved coding and knowledge‑work scores on Google‑reported benchmarks.
  • Gemini 3.5 Flash‑Lite targets high‑throughput, low‑latency tasks by delivering roughly 350 output tokens per second and a lower price point of $0.30 per million input tokens and $2.50 per million output tokens for large‑volume workloads.
  • Gemini 3.5 Flash Cyber is a cybersecurity‑tuned model designed to find and fix code vulnerabilities inside Google’s CodeMender agent and will be offered only to governments and vetted partners in a limited pilot to reduce dual‑use risk.
  • Gemini 3.5 Pro remains in partner testing with no broad release date and Google has begun pretraining runs for Gemini 4, a sequence that keeps pressure on enterprise customers weighing cost, coding performance, and choices between Google, Anthropic, and OpenAI.