Chrome Built-in AI is Chrome’s effort to let websites and web apps use browser-managed AI models and APIs. In Chrome, that includes Gemini Nano and task-specific APIs exposed to web pages.
Unlike sending every text task to a server, supported environments can run the work inside the browser. That can reduce latency, reduce server costs, and make it easier to keep user input on the device.
Different APIs serve different jobs
Chrome Built-in AI covers several task-focused APIs, including translation, language detection, summarization, writing, rewriting, and proofreading. Their availability differs: some are in Chrome Stable, some are in origin or developer trials, and some are limited to early preview access. Registration status and runtime requirements can also change.
- Translator API: translate text
- Language Detector API: detect input language
- Summarizer API: condense long content
- Writer API: create text for a specified task
- Rewriter API: revise existing text length or tone
- Proofreader API: assist with readability and grammar
Design around support checks
Built-in AI is not available in every Chrome environment. Support depends on the API, Chrome version, operating system, available storage, memory, GPU or CPU, and whether the model has already been downloaded. Some APIs do not support mobile Chrome yet.
A practical implementation flow
- Feature-detect the API you want to use
- Call availability() to check the current state
- Show progress when the model is downloadable
- Provide a clear unsupported state
- Review AI output before treating it as final
Features that depend on trial-stage APIs, such as the Rewriter API, should not be treated as generally available until the current registration status, target Chrome versions, and user environment requirements are clear.