3.6 Flash for general production work
3.6 Flash is described as the workhorse model in the release, with stronger coding, knowledge work, multimodal performance, and improved computer use over 3.5 Flash.
Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber are Google Gemini models for production AI agents, built for efficiency, low latency, computer use, and security workflows.
Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber are new additions to Google’s Flash model family, introduced for developers and customers building production AI agents. The release frames them around three needs: higher token efficiency, lower latency, and more reliable performance at scale.
3.6 Flash is positioned as the main workhorse model for coding, knowledge work, multimodal tasks, and computer use. 3.5 Flash-Lite targets the fastest, lowest-cost end of the family for high-throughput workflows such as agentic search and document processing. 3.5 Flash Cyber is presented as a specialized cyber-focused model used with CodeMender for security applications.
3.6 Flash is described as the workhorse model in the release, with stronger coding, knowledge work, multimodal performance, and improved computer use over 3.5 Flash.
The release says 3.6 Flash uses fewer output tokens, fewer reasoning steps, and fewer tool calls than 3.5 Flash on multi-step workflows, which lowers cost and latency.
3.5 Flash-Lite is positioned as the fastest 3.5-class model, with a throughput figure of 350 output tokens per second in the cited index.
3.5 Flash-Lite includes built-in computer use support and can be tuned with minimal, low, or higher thinking levels depending on workload needs.
3.5 Flash Cyber is paired with CodeMender for cybersecurity work, where careful orchestration between model and agent infrastructure is required.
The models are presented as tools for production AI agents, with emphasis on reliability, efficiency, and scale rather than single-turn chat.
Use 3.6 Flash when the task requires stronger coding, document understanding, or multimodal analysis, such as parsing charts, reviewing reports, or drafting work products.
Use 3.6 Flash or 3.5 Flash-Lite when an agent must take actions across browser, mobile, or desktop interfaces, especially for long-horizon workflows and enterprise automation.
Use 3.5 Flash-Lite for large volumes of search, classification, extraction, or document processing where latency and throughput matter more than heavyweight reasoning.
Use 3.5 Flash-Lite with higher thinking levels when a smaller model still needs to coordinate subagent-style or multi-step tasks without giving up speed.
Use 3.5 Flash Cyber with CodeMender for security-related work that depends on model plus agent orchestration, such as finding and fixing vulnerabilities efficiently.
The source describes Gemini 3.6 Flash as a workhorse model for better coding, knowledge work, multimodal tasks, and computer use. Gemini 3.5 Flash-Lite is positioned for low-latency, high-throughput workflows such as agentic search and document processing. Gemini 3.5 Flash Cyber is presented as a specialized cyber-focused model paired with CodeMender for security work.
Yes. The source says computer use is built into Gemini 3.6 Flash and 3.5 Flash-Lite, and it is also integrated natively into Gemini 3.5 Flash in the earlier computer-use announcement. Developers can use it for agents that act across browser, mobile, desktop, and client-side surfaces.
The source says developers and enterprises can access computer use in Gemini 3.5 Flash via the Gemini API and Gemini Enterprise Agent Platform. It also mentions computer use as a built-in client-side tool via the Gemini API and Gemini Enterprise for Gemini 3.6 Flash.
No pricing page details were available in the provided evidence. The launch post itself does, however, give API pricing for 3.6 Flash at $1.50 per 1M input tokens and $7.50 per 1M output tokens, and for 3.5 Flash-Lite at $0.30 per 1M input tokens and $2.50 per 1M output tokens.
The source highlights enterprise and production use cases, including continuous software testing, knowledge work, document parsing, chart and data analysis, report drafting, agentic search, and high-volume document processing.
AakarDev AI helps teams manage AI provider access, project setups, logs, and analytics in one dashboard with BYOK support.
Orca is an Agent Development Environment for shipping with coding agents, running multiple CLI agents in parallel across isolated worktrees, with desktop and mobile workflows.
AI Magicx is a unified AI workspace for chat, image, video, voice, music, email and developer tasks, helping teams and creators manage multiple models in one place.
Paper is a design tool that connects canvas, code, and AI agents so teams can create, share, and ship work in one workflow. Includes desktop app and MCP access.
blop is a QA agent that writes browser tests as code in your repo, runs them in CI, clusters repeated failures, and can open PRs to fix broken tests.
RLAMA is a local AI platform for building RAG systems and intelligent agents on macOS, Linux, and Windows, with HTTP API support.