Google Antigravity

Agent-first

The agent-first surface built to orchestrate multiple subagents across all your environments. This feature allows users to run complex, long-running workflows in the background while other agents handle everything from research to implementation to testing.

Antigravity 2.0

Your command center to manage multiple local agents in parallel. Users can group conversations into Projects, operate across multiple workspaces, and automate routine tasks with scheduled messages, enhancing productivity and organization.

Antigravity CLI

The lightweight, fast, terminal-first surface to work with Antigravity agents. Users can run autonomous coding agents, execute shell commands directly, and manage background subagents all from their keyboard, streamlining their workflow.

Antigravity IDE

The fully-featured, agentic IDE comes complete with the agent manager, artifacts, and a deep understanding of your codebase. This allows users to develop and manage their projects efficiently within a single environment.

Antigravity SDK

Prototype custom agents leveraging Antigravity’s harness with minimal code. Users can create simple Python scripts to iterate on agentic applications, automate software engineering tasks, and run evaluations on top of the Antigravity agent harness.

Automated Scheduled Task and Skills

Deploy scheduled cron agents to run routine tasks like auditing sites, scanning documents for errors, or generating daily briefs. Users can create custom skills to automate workflows, significantly reducing manual effort.

Parallel Agent Execution

Offload research, sourcing, and drafting to specialized subagents running in parallel to accelerate work. This feature allows users to manage multiple tasks simultaneously, enhancing productivity and efficiency.

Guidance to go from 90% to 100%

Intuitively provide feedback across every surface and Artifact to steer the agent to your desired outcomes. This feature helps users refine their projects and achieve optimal results.

Verifiable Scientific Artifacts

Build trust with results that are grounded in evidence and legible at each step. This feature ensures that users can rely on the outputs generated by the agents, enhancing credibility.

Artifacts and Verification

Agents prioritize verifying information and interacting with users through artifacts and tasks to build trust in their code. This feature is essential for users who need to ensure the reliability of their outputs.

Specialized Science Skills

Give your agents fluency in the tools, models, and databases that researchers use—from frontier models like AlphaGenome to over 30 major scientific databases including AlphaFold Database UniProt, PubChem, and ChEMBL. This feature enhances the agents' capabilities in scientific research.

Browser Use

The Agent is able to use Chrome to navigate to pages and interact with them, enabling browser-in-the-loop testing of its own code. This feature allows users to test and validate their code in real-time within the browser environment.

Visual Artifacts

The Agent is able to take screenshots and videos of its work in the browser so that users can see the results of the work asynchronously. This feature provides visual feedback and documentation of the agents' activities.

Visual Feedback

Users are able to leave visual comments on the visual Artifacts, making it simple to provide specific targeted feedback to the Agent. This feature enhances collaboration and improves the quality of the output.