chatboxai/chatbox · error · ApiError

Provider doesnt support image generation

Error message

Provider doesnt support image generation

What it means

Thrown by AbstractAISDK.paint when this.getImageModel() returns a falsy value — the current provider configuration exposes no image-generation model to the AI SDK. The check happens before generateImage is called, so no provider request is attempted; this is a capability-mismatch error, not a network error.

Source

Thrown at src/shared/models/abstract-ai-sdk.ts:431

        () => {}
      )
      await streamIterator.return?.().catch(() => {})
    }
  }

  public async paint(
    params: {
      prompt: string
      images?: { imageUrl: string }[]
      num: number
      aspectRatio?: string
    },
    signal?: AbortSignal,
    callback?: (picBase64: string) => void | Promise<void>
  ): Promise<string[]> {
    const imageModel = this.getImageModel()
    if (!imageModel) {
      throw new ApiError('Provider doesnt support image generation')
    }
    const result = await generateImage({
      model: imageModel,
      prompt: params.prompt,
      // images 暂时不支持
      n: params.num,
      abortSignal: signal,
      // Image generation is billable; network-error retries could double-charge.
      maxRetries: 0,
    })
    const dataUrls = result.images.map((image) => `data:${image.mediaType};base64,${image.base64}`)
    for (const dataUrl of dataUrls) {
      await callback?.(dataUrl)
    }
    return dataUrls
  }

  /**

View on GitHub (pinned to 81571269ad)

Solutions

1. Switch to a provider+model that supports image generation (e.g. OpenAI gpt-image, a DALL·E model, Google Imagen) before invoking paint.
2. If you control the provider config, populate the imageModel field / implement getImageModel() to return a valid SDK image model.
3. In the UI, gate the 'generate image' control on whether the active provider exposes an image model (call getImageModel() to enable/disable the button).
4. Catch this ApiError and show the user a targeted 'this model cannot generate images' message rather than a generic failure.

Example fix

// before
const imageModel = this.getImageModel()
if (!imageModel) {
  throw new ApiError('Provider doesnt support image generation')
}
// after — include the provider/model id for actionable UX
const imageModel = this.getImageModel()
if (!imageModel) {
  throw new ApiError(`Provider ${this.name} (${this.options.model.modelId}) does not support image generation`)
}

// Before calling paint(), confirm the provider exposes an image model.
function providerCanPaint(sdk: AbstractAISDK): boolean {
  return Boolean(sdk.getImageModel())
}
if (!providerCanPaint(sdk)) {
  showUser('This model cannot generate images. Pick an image-capable model.')
  return
}

function supportsImageGeneration(sdk: { getImageModel(): unknown }): boolean {
  return Boolean(sdk.getImageModel())
}

try {
  await sdk.paint({ prompt, num })
} catch (e) {
  if (e instanceof ApiError && /doesn't support image generation/i.test(e.message)) {
    promptUserToSelectImageModel()
    return
  }
  throw e
}

When it happens

Trigger: paint() is invoked (image generation requested) on a provider whose getImageModel() returns undefined — e.g. a chat-only provider, a provider whose config doesn't declare an image model, or a provider type for which image generation isn't implemented in this SDK subclass.

Common situations: User picked a text-only model (e.g. a generic OpenAI chat model) and clicked 'generate image'; provider config missing the imageModel field; model id resolves to a chat model but the UI offered the paint action; a custom provider that doesn't expose DALL·E/Imagen-style endpoints.

Related errors

• No readable text content found in EPUB file
• Embedding batch failed: expected ${batchTexts.length}, got $
• Custom provider "${providerId}" conflicts with a builtin pro
• Image generation requires an active chat session.
• Missing --prompt.

AI-assisted analysis of chatboxai/chatbox@81571269ad (2026-08-12). Data as JSON: /api/errors/0bf617690263ec81. Report an issue: GitHub.