Home/AI NOW/Google launches Gemini 3.6 Flash and 3.5 Flash-Lite for more efficient agents

ModelsNEW MODEL

Google launches Gemini 3.6 Flash and 3.5 Flash-Lite for more efficient agents

Google has released Gemini 3.6 Flash with improved coding, knowledge-work, multimodal and token efficiency, alongside 3.5 Flash-Lite for high-throughput workloads.

07/22/20261 sources reviewed

Quick summary

Google says 3.6 Flash uses 17% fewer output tokens than 3.5 Flash on the Artificial Analysis Index.
3.6 Flash costs $1.50 per million input tokens and $7.50 per million output tokens.
3.5 Flash-Lite advertises 350 output tokens per second and pricing of $0.30 input and $2.50 output per million tokens.

WHAT HAPPENED

What happened?

Google has released Gemini 3.6 Flash with improved coding, knowledge-work, multimodal and token efficiency, alongside 3.5 Flash-Lite for high-throughput workloads.

The update makes workload-specific model routing more important than choosing one flagship for everything. Teams could reserve 3.6 Flash for complex planning and use Flash-Lite for repetitive extraction or classification, but should measure total cost with their own prompt lengths and retry rates.

WHY IT MATTERS

Why does it matter?

Agent operating cost depends not only on per-call pricing but also on output tokens, tool calls and throughput, so both models can directly change the economics of large-scale automation.

WHO SHOULD CARE

Who should care?

AI developersAgent operations teamsEnterprises

RELATED AI

Related AI

AIZIGOO VIEW

AIZIGOO view

Speed and benchmark figures come from Google and the cited evaluators; real-world results may vary by workload.