Gemini 3.6 Flash

Changes from previous version

Higher precision and token efficiency than Gemini 3.5 Flash on coding, computer use, and knowledge work at lower per-output-token cost.

Release Summary

Google's agentic workhorse Flash model after Gemini 3.5 Flash: 17% fewer output tokens on the Artificial Analysis Index, with gains on DeepSWE (49%), MLE Bench (63.9%), OSWorld-Verified (83.0%), and GDPval-AA v2 (1421). Priced at $1.50/M input and $7.50/M output, below 3.5 Flash output rates.

Timeline

Other Gemini versions