nexu-io/open-design · error · Error

${tag} ${resp.status}: ${truncate(text, 240)}

Error message

${tag} ${resp.status}: ${truncate(text, 240)}

What it means

Thrown by renderOpenAIImage after the images/generations fetch returns a non-2xx status. The tag is 'azure-openai' when detectAzureEndpoint flagged the base URL, otherwise 'openai'. The message includes resp.status and the first 240 chars of the response body so quota/auth/model errors are visible. This is the catch-all upstream HTTP failure for the OpenAI image path.

Source

Thrown at apps/daemon/src/media/index.ts:953

    // Azure's canonical auth header. Some deployments accept Bearer
    // (the curl example we tested against does) but api-key is what
    // their docs document, so send both. OpenAI ignores unknown
    // headers, so this is harmless on the standard endpoint too.
    headers['api-key'] = credentials.apiKey;
  }

  const resp = await fetch(url, withMediaRequestInit(ctx, {
    method: 'POST',
    headers,
    body: JSON.stringify(body),
    dispatcher: ctx.requestInit.dispatcher
      ?? openAIImageDispatcher as unknown as NonNullable<RequestInit['dispatcher']>,
    signal: AbortSignal.timeout(Math.max(OPENAI_IMAGE_HEADERS_TIMEOUT_MS, OPENAI_IMAGE_BODY_TIMEOUT_MS)),
  }));
  const text = await resp.text();
  if (!resp.ok) {
    const tag = azure ? 'azure-openai' : 'openai';
    throw new Error(`${tag} ${resp.status}: ${truncate(text, 240)}`);
  }
  let data: any;
  try {
    data = JSON.parse(text);
  } catch {
    throw new Error(`openai non-JSON response: ${truncate(text, 200)}`);
  }
  const entry = data && Array.isArray(data.data) ? data.data[0] : null;
  if (!entry) throw new Error('openai response had no data[0]');
  let bytes;
  if (entry.b64_json) {
    bytes = Buffer.from(entry.b64_json, 'base64');
  } else if (entry.url) {
    const imgResp = await fetch(entry.url, withMediaRequestInit(ctx));
    if (!imgResp.ok) throw new Error(`openai image fetch ${imgResp.status}`);
    const arr = await imgResp.arrayBuffer();
    bytes = Buffer.from(arr);
  } else {

View on GitHub (pinned to 5be4028344)

Solutions

  1. Read the embedded status and body snippet: 401/403 -> rotate/regenerate the key; 429 -> slow down or upgrade plan; 404 -> verify model id and Azure deployment name.
  2. Confirm credentials.baseUrl is correct (omit it for public OpenAI; set the full Azure deployment URL for Azure).
  3. Retry after the provider's Retry-Backoff for 429/5xx; the OpenAI image path does not auto-retry.
  4. Check the provider status page if 5xx persists.

Example fix

// before: wrong Azure deployment name -> 404
OD_OPENAI_BASE_URL=https://my-resource.openai.azure.com/openai/deployments/wrong-name
// after
OD_OPENAI_BASE_URL=https://my-resource.openai.azure.com/openai/deployments/correct-deployment
Defensive patterns

Strategy: retry

Validate before calling

function classifyOpenAIStatus(status: number): 'retry'|'fatal'|'config' {
  if (status === 429 || status >= 500) return 'retry';
  if (status === 401 || status === 403) return 'config';
  return 'fatal';
}

Try / catch

try {
  return await renderOpenAIImage(ctx, credentials);
} catch (err) {
  const m = err instanceof Error ? err.message : '';
  const statusMatch = m.match(/(?:azure-openai|openai) (\d{3}):/);
  const status = statusMatch ? Number(statusMatch[1]) : 0;
  if (status === 429 || status >= 500) throw new RetryableError(m);
  if (status === 401 || status === 403) throw new ConfigError('rotate OpenAI key');
  throw err;
}

Prevention

When it happens

Trigger: Invalid or revoked API key (401); exhausted quota or billing issue (402/429); requesting a model the deployment does not expose (404 on Azure); rate limiting (429); transient 5xx from OpenAI/Azure; wrong baseUrl pointing at a non-OpenAI host.

Common situations: Free-tier key hitting rate limits; Azure deployment name mismatch; baseUrl typo; shared key revoked; model deprecated by provider.

Related errors


AI-assisted analysis of nexu-io/open-design@5be4028344 (2026-08-12). Data as JSON: /api/errors/a3be5cf78d8091e8. Report an issue: GitHub.