πŸ”„ Content Genius Generation Endpoint Migration

Camilo Aponte

Camilo Aponte

Last updated on Nov 5, 2026

πŸ“Œ What changed

Content Genius still supports article generation. For current v5 API integrations, use POST /api/cgenius/v5/specialized/generate/. This operation can create and start a new article; existing articles have separate retry and regeneration workflows. Confirm the exact older endpoint before describing it as retired. Source: content-assistant/app/cgenius/v5/api/views/articles.py:3–16, 279–295.

Use the endpoint labeled for starting or generating content on an existing record in the in-platform API reference. Do not reuse saved URLs, collection files, or code from before the migration. The reference in your account is the source of truth for the path, version, and required fields.

πŸš€ Start full generation

You can create and start a Content Genius article directly through the generation API. Supply the topic and required Brand Vault context. Use the appropriate retry or regeneration workflow when working with an existing article. Source: content-assistant/app/cgenius/v5/api/views/articles.py:221–237, 298–355.

  1. Consult the Content Assistant API documentation and locate the Content Genius v5 specialized generation operations. Confirm the service host and authentication requirements before making requests. Source: content-assistant/app/app/urls.py:34–41, 58–60.
  2. For a new article, use POST /api/cgenius/v5/specialized/generate/. To retry an existing eligible article, use POST /api/cgenius/v5/specialized/articles/{article_id}/retry/. Follow the documentation for the specific operation, and avoid operations explicitly marked deprecated. Source: content-assistant/app/cgenius/v5/api/urls.py:99, 162.
  3. Send that documented method, normally a write request, to the documented path only. Include your current API credentials the same way the reference shows.
  4. For new generation, send query and the appropriate brand_vault_uuid, together with any supported optional fields such as article_type, article_length, keywords, language_code, and request_id. No full-generation setting or existing article identifier is required. Send only inputs supported by the selected operation. Source: content-assistant/app/cgenius/v5/api/views/articles.py:221–295.
  5. Do not add unofficial version prefixes or swap the record identifier into an old path.

A 202 response does not mean the article is finished. Store the returned article_id and inspect status. A newly started generation enters planning before progressing through generation; the response also provides chat_id and ws_url. Reusing an idempotency request_id may return an existing article and its current state. Source: content-assistant/app/cgenius/v5/api/views/articles.py:333–355, 468–474, 560–578.

βœ… Check until completion

Generation runs asynchronously. Poll GET /api/cgenius/v5/specialized/articles/{article_id}/ for the returned article, or use the returned ws_url to watch its chat updates. The status endpoint also supports ?view=status for a compact response. Source: content-assistant/app/cgenius/v5/api/views/articles.py:16, 1348–1355.

  • Planning, queued, pending, and generating are nonterminal article states. Continue monitoring the existing article rather than starting a duplicate generation. Use a consistent request_id when safely repeating the original create request. Source: content-assistant/app/cgenius/v5/models.py:311–336.
  • When status is generated, generation has finished. Retrieve the full article response from GET /api/cgenius/v5/specialized/articles/{article_id}/ without the compact status projection. Review ship_status and verification results before publishing; generated content may still need review or revision. Source: content-assistant/app/cgenius/v5/models.py:302–345; content-assistant/app/cgenius/v5/api/views/articles.py:1348–1354.
  • When status is failed, inspect error and failure_code in the status response. Resolve any actionable issue, such as quota exhaustion, and use POST /api/cgenius/v5/specialized/articles/{article_id}/retry/ when the article is eligible. Some failures are transient; others require changes before retrying. Source: content-assistant/app/cgenius/v5/api/views/articles.py:736–821, 1361–1381.

If a request returns not found, method not allowed, or a version-related error, verify the host, path, HTTP method, identifier, authentication scope, and response details. Compare the request with the current API documentation. Do not assume the endpoint is retired solely from the error status. Source: content-assistant/app/cgenius/v5/api/views/articles.py:750–756, 1344–1355.

πŸ› οΈ If it still fails

Confirm that the record exists in the same account as the API credentials, that billing and permissions allow Content Genius generation, and that the body matches the current schema with no extra legacy fields.

If you need further assistance, open the chat widget in the bottom-right corner of the platform and send your message. A member of our support team will reply in the chat.

Was this article helpful?