Gemini 3.6 Flash
Google DeepMind🇺🇸 USFrontier· Jul 21, 2026
TextReasoningVisionAPI
Gemini 3.6 Flash consumes 17 % fewer output tokens than its predecessor and completes multi-step workflows with fewer reasoning steps and tool calls. It targets coding, knowledge work, and agentic automation at lower cost.
Summary of the model card
- Context
- 1M tokens
- Price / 1M tok
- $0.38 in · $1.88 out
- Modalities
- text, image, video, document, audio → text
2 variants
Sources
OpenRouterBlogDeutsch
Das Modell verbraucht 17 % weniger Ausgabetokens als der Vorgaenger und bearbeitet Mehrschrittaufgaeben mit weniger推理-Schritten und Tool Calls. Es richtet sich an Coding, Wissensarbeit und agente Automatisierung zu niedrigeren Kosten.
Provenance
- Detected
- Sep 27, 2026
- First source
- openrouter
- Description
- Summary of the model card
- Editorially reviewed
- —