v137 · miscellaneous

Writer API

The Writer API can be used for writing new material given a writing task prompt, backed by an on-device AI language model. Developers will be able to use this API to generate textual explanations of structured data, composing a post about a product based on reviews or product description, expanding on pro and con lists into full views and more. An enterprise

concepts

  1. Write

    On-device Writer — generate prose from a brief without a server round-trip. Pairs with Rewriter for full author/edit pipelines on-device.

  2. Data to Prose

    The explainer's "generate textual explanations of structured data" use case. Paste JSON (review, meeting notes, incident metadata) and watch the Writer expand it into a paragraph.

  3. Recipe book

    Four production-ready recipes: release notes, product description, meeting follow-up, weekly status. Each card sets shared context, tone, length, and format, and ships with seed text + a deterministic fallback when the API isn’t available.

  4. Streaming + per-call context

    One Writer with a brand-voice shared context, three parallel streams — each adding a different per-call audience context. Demonstrates writeStreaming plus the context-reuse pattern that keeps repeat calls cheap.

why it shipped

Browsers and operating systems are increasingly expected to gain access to a language model. By exposing this built-in model, we avoid every website needing to download their own multi-gigabyte language model, or send input text to third-party APIs. The writer API in particular exposes a high-level API for interfacing with a language model in order to produce new output for a variety of use cases [1], in a way that does not depend on the specific language model in question.

references