Warp
@warpdotdev
The Warp Agent now supports Gemini 3.5 Flash. It's an impressive combination of intelligence, speed, and cost. https://t.co/6s1YCVpw7m
3retweets59likes
View on XWarp Adds Gemini 3.5 Flash to Power High Speed Agentic Terminal Workflows
Warp, an agentic development environment that combines a terminal and code editor, added support for
Gemini 3.5 Flash to its autonomous agent. This integration allows developers to use Google's latest high-speed model for agentic tasks, following the platform's recent launch of multi-agent orchestration in Oz.This update arrives alongside a broader industry shift toward Google's Gemini 3.5 Flash release, as developers prioritize inference speed and cost for autonomous loops. While flagship models offer deeper reasoning, the Flash series is specifically tuned for the high-token, multi-turn interactions required when agents navigate complex codebases.
You can now select Gemini 3.5 Flash within the Warp Agent settings to power your local development workflows. This model choice complements Warp's open-weight model routing, giving you more granular control over the balance between intelligence and token expenses during long-running agent sessions.
Still wondering? A few quick answers below.
The Warp Agent is an autonomous AI component integrated directly into the Warp terminal and code editor. It is designed to handle multi-step software development tasks by navigating codebases, running terminal commands, and executing tests independently. This allows developers to delegate complex engineering workflows to an agent that reasons through problems in a continuous loop.
Integrating Gemini 3.5 Flash provides the Warp Agent with a high-speed and cost-effective model option for autonomous workflows. Because agentic tasks require frequent turns and high token usage, this model optimizes performance by reducing latency during multi-step interactions. It offers a balance of intelligence and execution speed specifically tuned for the demands of software engineering.
Warp has officially added support for Gemini 3.5 Flash within its agentic development environment. Users can now select this model as their preferred engine for terminal-based agent tasks through the platform settings. This update gives developers more flexibility to choose between different frontier models based on their specific requirements for reasoning depth, speed, and token expenses.
Speed is critical for agentic coding because autonomous agents operate in iterative loops that involve observing the environment, reasoning, and taking action. High-speed models like Gemini 3.5 Flash minimize the time spent waiting for the AI to process each step. This lower latency makes long-running development tasks, such as site migrations or code verification, more practical for daily use.
