Prompt API

at a glance

why it exists

Cloud AI APIs require a network connection, a billing account, and introduce latency and privacy concerns for sensitive content. Devices running Chrome already download a small on-device LLM (Gemini Nano) for browser-level features. The Prompt API exposes that same model to web pages through a sandboxed JS interface, letting developers add language model capabilities without external dependencies, API keys, or sending user data off-device.

shape of the API

All interactions flow through a session object. The static LanguageModel namespace creates and introspects sessions; each session maintains a rolling conversation context and token budget.

LanguageModel static methods

LanguageModelSession instance members

recipes

Check model availability and create an active session with LanguageModel.create():

const avail = await LanguageModel.availability();
if (avail === 'unavailable') {
  console.warn('Prompt API not supported on this device');
} else {
  // 'downloadable' → browser will fetch the model on create()
  const session = await LanguageModel.create({
    initialPrompts: [
      { role: 'system', content: 'You are a concise technical writer.' }
    ]
  });
  const answer = await session.prompt('Explain CSS cascade in one sentence.');
  console.log(answer);
  session.destroy();
}

Streaming response

const session = await LanguageModel.create();
const stream = session.promptStreaming('Write a haiku about the browser cache.');
let text = '';
for await (const chunk of stream) {
  text += chunk;
  outputEl.textContent = text;
}
session.destroy();

Structured output (JSON schema constraint)

const session = await LanguageModel.create();
const schema = {
  type: 'object',
  properties: {
    sentiment: { type: 'string', enum: ['positive', 'neutral', 'negative'] },
    score: { type: 'number', minimum: 0, maximum: 1 }
  },
  required: ['sentiment', 'score']
};
const raw = await session.prompt(
  'Analyse: "The docs were clear and the API is elegant."',
  { responseConstraint: schema }
);
const { sentiment, score } = JSON.parse(raw);
session.destroy();

Multimodal input (image)

const session = await LanguageModel.create({
  expectedInputs: [{ type: 'image' }]
});
const blob = await fetch('/screenshot.png').then(r => r.blob());
const result = await session.prompt([
  { type: 'text', value: 'Describe what is wrong in this UI screenshot.' },
  { type: 'image', value: blob }
]);
session.destroy();

Session lifecycle and context management

// Standard-web-page example. Note: the README marks the temperature/topK create
// options deprecated for standard web contexts (extension contexts only); they are
// shown because the syntax table above documents them as part of the shipped 148
// surface. LanguageModel.params() is extension-only and is NOT called here (see below).
const session = await LanguageModel.create({
  temperature: 0.7,
  topK: 3,
  initialPrompts: [{ role: 'system', content: 'Answer concisely.' }]
});

// Seed few-shot context without generating a reply
await session.append([{ role: 'user', content: [{ type: 'text', value: 'Example Q&A pair for style.' }] }]);

// Budget check: tokens a hypothetical prompt would consume vs the session budget
const projected = await session.measureContextUsage(largeDocument);
console.log(projected, session.contextUsage, 'of', session.contextWindow);

// clone() copies the session including conversation history
const branch = await session.clone();
branch.destroy();
session.destroy();

// Extension contexts only: LanguageModel.params() reports the model's sampling-parameter
// defaults and maxima; per the README it is not functional in standard web pages.
// const limits = await LanguageModel.params();
// -> { defaultTemperature, maxTemperature, defaultTopK, maxTopK }

browser support

see also

• webmachinelearning/prompt-api — Explainer and spec repo
• Prompt API specification draft
• chromestatus — Prompt API (5134603979063296)
• chromestatus — API record 5134603979063296
• chromestatus — milestone=148 listing
• Chrome for Developers — Prompt API guide
• W3C TAG design-reviews #1093
• Chromium Issue 417526788
• BCD — api/LanguageModel.json
• chrome-platform-showcase — live demo