vercel/ai · error · Error

Amazon Bedrock returned no images. Status: ${response.status

Error message

Amazon Bedrock returned no images. Status: ${response.status}

What it means

After the moderation check, the provider verifies the Nova Canvas response actually contains images. If response.images is missing or empty (for any status other than moderation), doGenerate throws 'Amazon Bedrock returned no images' with the response status appended. This guards against silent empty results from the provider.

Source

Thrown at packages/amazon-bedrock/src/amazon-bedrock-image-model.ts:286

      ),
      abortSignal,
      fetch: this.config.fetch,
    });

    // Handle moderated/blocked requests
    if (response.status === 'Request Moderated') {
      const moderationReasons = response.details?.['Moderation Reasons'];
      const reasons = Array.isArray(moderationReasons)
        ? moderationReasons
        : ['Unknown'];
      throw new Error(
        `Amazon Bedrock request was moderated: ${reasons.join(', ')}`,
      );
    }

    // Check if images are present
    if (!response.images || response.images.length === 0) {
      throw new Error(
        'Amazon Bedrock returned no images. ' +
          (response.status ? `Status: ${response.status}` : ''),
      );
    }

    return {
      images: response.images,
      warnings,
      response: {
        timestamp: currentDate,
        modelId: this.modelId,
        headers: responseHeaders,
      },
    };
  }
}

function getBase64Data(file: ImageModelV4File): string {

View on GitHub (pinned to 69428b1f8b)

Solutions

Update @ai-sdk/amazon-bedrock to the latest version
Validate your generation config (width/height/quality/count) against the Nova Canvas limits for your model ID
Simplify the prompt and retry to rule out soft filtering
Log/report the response.status from the error message; open an issue with AWS if a valid request consistently yields no images

Example fix

// before
await generateImage({ model, prompt: 'a cat', providerOptions: { bedrock: { imageGenerationConfig: { width: 99999, height: 99999 } } } });
// after
await generateImage({ model, prompt: 'a cat', providerOptions: { bedrock: { imageGenerationConfig: { width: 1024, height: 1024 } } } });

Defensive patterns

Strategy: retry

Validate before calling

// validate generation config against model limits, e.g. 320–4096 px for Nova Canvas
guard(width >= 320 && width <= 4096 && height >= 320 && height <= 4096)

Try / catch

try {
  const { images } = await generateImage({ model, prompt });
} catch (e) {
  if (e instanceof Error && e.message.includes('returned no images')) {
    // adjust prompt/config and retry
  }
}

Prevention

Validate width/height/quality/count against Nova Canvas limits
Retry with a simplified prompt when generation returns empty
Update the provider package for newer model response shapes

When it happens

Trigger: generateImage against a bedrock image model where the API responds successfully but with zero images — e.g. unacknowledged filtering, invalid generation config (width/height out of range), or quality/seed issues.

Common situations: Invalid imageGenerationConfig dimensions for the model; prompt rejected softly by the model; model returning an unusual status the package doesn't explicitly map; new Nova model versions with different response shapes on outdated provider packages.

Related errors

AI-assisted analysis of vercel/ai@69428b1f8b (2026-08-30). Data as JSON: /api/errors/7897e3f26fabfc51. Report an issue: GitHub.