New forum post submission
AI Models & Assistants / Knowledge library
text-moderation
Previous generation text-only moderation model
Why learn text-moderation?
Offers the documented OpenAI model capabilities through hosted.
The problem it solves
Provides model inference for supported text inputs.
Where it came from
text-moderation: Previous generation text-only moderation model Published aliases: text-moderation-007. Input: text. Output: text. Checked 2026-10-13. Model capabilities are distinct from tools available in a particular agent.
Build your understanding
Work through these exercises in order. Try the task before opening a reference.
- Explain what text-moderation does.
- Identify a use case for text-moderation.
- Describe a limitation of text-moderation.
- Name a compatible tool for text-moderation.
- Compare text-moderation against an alternative.
Documentation and sources
- Primary source / model card (developers.openai.com)
Catalog research notes
- See the AI Workbench for source dates, pricing evidence and configuration details.
Connect the ideas
Built on
Real World Example
Community Forum Safety Enforcement
A regional hobbyist forum platform that hosts user-generated discussion threads.
The business problem
The platform faces high operational costs and inconsistent enforcement when human moderators manually review thousands of daily posts for harassment or hate speech.
How this topic solves it
The platform integrates the text-moderation-latest model via the /v1/moderations endpoint to automatically flag incoming user posts against safety categories before they are published to the public feed.
- Configure the application backend to intercept new post submissions before database persistence.
- Send the raw text content of the post to the /v1/moderations endpoint.
- Parse the JSON response to check the boolean flags for categories like hate, harassment, or self-harm.
- Route posts flagged as true to a human review queue and allow non-flagged posts to proceed to the public feed.
The flow of a user post through the automated moderation system.
text-moderation-latest endpoint
Flagged for manual review
Post published to site
- User Input → sends text → Moderation API
- Moderation API → if flagged → Human Queue
- Moderation API → if safe → Public Feed
Expected business outcome
The platform achieves a consistent baseline for content safety, significantly reducing the volume of benign content that human moderators must manually inspect.
Limitations and checks
The model is a text-only moderation tool, meaning it cannot analyze images or embedded media attached to posts; developers must implement separate validation logic for non-text content.