Google’s Gemini suite has just released two GA models—Gemini 3.6 Flash and Gemini 3.5 Flash‑Lite—ready for production use. The 3.6 Flash model delivers stronger performance on complex agentic and multimodal tasks while cutting token usage, and it comes at a lower price than the older 3.5 Flash. The 3.5 Flash‑Lite is the fastest, lowest‑cost model in the 3.5 family, outperforming earlier Flash‑Lite generations for high‑throughput execution.

Pricing is $1.50 per 1M input tokens and $7.50 per 1M output tokens for 3.6 Flash, while 3.5 Flash‑Lite costs $0.30 and $2.50. Both models support a 1M token context window, 64k max output, and the full suite of built‑in tools, balancing speed, intelligence, and cost for agentic workflows, code generation, and large‑scale data parsing.

New capabilities include token and turn reduction, higher‑quality code generation, improved multimodal reasoning on charts and images, and native Computer Use support. The models also feature better instruction following and an upfront programmatic inspection step that can add exploratory steps on simpler tasks.

API changes require removing deprecated sampling parameters (temperature, top_p, top_k) and disallowing prefilled model turns. Migrating involves applying the gemini‑interactions‑api skill and updating your code to the new models, ensuring smooth transition to the updated Gemini architecture.