linshenkx/prompt-optimizer · error · RequestConfigError

Chrome built-in AI model is not downloaded. Open the model m

Error message

Chrome built-in AI model is not downloaded. Open the model manager and click download to prepare it.

What it means

createReadySession in the Chrome built-in AI adapter checks window.ai availability; when the underlying model reports 'downloadable', it means the Gemini Nano weights are not yet on the device and the adapter throws a RequestConfigError telling the user to download the model first.

Source

Thrown at packages/core/src/services/llm/adapters/chrome-built-in-adapter.ts:112

        }
      })
    } finally {
      session.destroy?.()
    }
  }

  protected getParameterDefinitions(_modelId: string): readonly ParameterDefinition[] {
    return []
  }

  protected getDefaultParameterValues(_modelId: string): Record<string, unknown> {
    return {}
  }

  private async createReadySession(initialPrompts?: ChromeLanguageModelPrompt[]) {
    const status = await checkChromeBuiltInAvailability()
    if (status.availability === 'downloadable') {
      throw new RequestConfigError('Chrome built-in AI model is not downloaded. Open the model manager and click download to prepare it.')
    }
    if (status.availability === 'downloading') {
      throw new RequestConfigError('Chrome built-in AI model is still downloading. Please wait for the download to finish.')
    }
    if (status.availability !== 'available') {
      throw new RequestConfigError(status.error || 'Chrome built-in AI is not available in this browser.')
    }

    return await createChromeBuiltInSession(initialPrompts?.length ? { initialPrompts } : {})
  }

  private buildPrompt(messages: Message[]): {
    initialPrompts?: ChromeLanguageModelPrompt[]
    prompt: ChromeLanguageModelPromptInput
  } {
    const normalizedMessages = messages
      .map((message) => ({
        role: message.role,

View on GitHub (pinned to 3e677b1d9f)

Solutions

  1. Open the model manager (or chrome://components → 'Optimization Guide On Device Model') and trigger the download, wait for it to finish
  2. Ensure Chrome is up to date and the Prompt API flag is enabled
  3. Free disk space and restart Chrome so the component can download
  4. Catch RequestConfigError in the UI and show the download prompt instead of failing silently

Example fix

// before
const s = await adapter.session()

// after
try {
  const s = await adapter.session()
} catch (e) {
  if (e instanceof RequestConfigError && /not downloaded/.test(e.message)) {
    showDownloadModelUI(); return
  }
  throw e
}
Defensive patterns

Strategy: validation

Validate before calling

import { checkChromeBuiltInAvailability } from '...'
const status = await checkChromeBuiltInAvailability()
if (status.availability === 'downloadable') {
  promptUserToDownloadModel() // instead of calling session()
}

Type guard

null

Try / catch

try { const s = await adapter.session() }
catch (e) {
  if (e instanceof RequestConfigError && /not downloaded/.test(e.message)) { promptDownload(); return }
  throw e
}

Prevention

When it happens

Trigger: Starting a chat session (session() → createReadySession) in a Chrome/preview browser where Prompt API is exposed but the model was never downloaded; fresh browser profile; model was evicted after disk cleanup.

Common situations: Enabling chrome://flags#prompt-api-for-gemini-nano without downloading the model, using a new profile, Chrome Dev/Canary where the component is optional, insufficient disk space preventing auto-download.

Related errors

AI-assisted analysis of linshenkx/prompt-optimizer@3e677b1d9f (2026-08-27). Data as JSON: /api/errors/b639f6f525bc3fbf. Report an issue: GitHub.