Google Integrates Anthropic Models into Antigravity AI Tool

Google has updated its Antigravity model picker to include Anthropic's Claude Opus 5.5 and Sonnet 5.5, marking a significant shift in the company's AI strategy. The addition of these models, which were previously available through separate subscriptions, allows developers to access both Google's Gemini 3.1 Pro and Anthropic's flagship coding models under one umbrella.

The move is notable given that Antigravity had been running older versions of Claude's Opus and Sonnet since last month, with the new models being added quietly without fanfare. This change now allows developers to run a live comparison between Gemini 3.1 Pro and Anthropic's 5.5 models every time they open a new session.

For teams deciding where their AI coding budget goes, this bundling changes the math. Paying Anthropic directly means being billed by the token through its API or through Cursor's wrapper around it, whereas paying Google comes with usage limits set by Google's tier rather than Anthropic's API pricing.

The strategy behind this move is not new for Google, which has a history of integrating rival models into its own products. By keeping Claude competitive inside Antigravity, developers have a reason to stay in Google's IDE even when they'd rather write Claude prompts than Gemini ones.