Content Moderation (Moderations)

OpenAI Moderations-compatible endpoint that classifies text for safety. The request body is the same as OpenAI's (input is required; model is optional and defaults to text-moderation-latest; omni-moderation-latest is recommended). The upstream response is returned as is.

Note: the gateway re-serializes the request body, so unknown top-level fields are dropped. Billed at the price of the corresponding model.

Source of truth in code: router/relay-router.go (httpRouter.POST("/moderations")), relay/helper/valid_request.go (required-input validation).

Current status (verified 2026-09-28): the route is enabled, but the model catalog currently contains no moderation models (both omni-moderation-latest and text-moderation-latest return 404 model_not_found). This endpoint becomes usable once a moderation model is listed.

Last verified/modified: 2026-09-28 (API Base URL unified to https://api.crazyrouter.com; example model names follow what GET /v1/models returns); verified 404 model_not_found, marked as in development for now

Provide your bearer token in the

Authorization

header when making requests to protected resources.

Example:

Authorization: Bearer ********************

curl --location 'https://api.crazyrouter.com/v1/moderations' \ --header 'Authorization: Bearer <token>' \ --header 'Content-Type: application/json' \ --data '{ "model": "omni-moderation-latest", "input": "I want to hurt someone." }'

{ "id": "modr-abc123", "model": "omni-moderation-latest", "results": [ { "flagged": true, "categories": { "violence": true, "hate": false }, "category_scores": { "violence": 0.91, "hate": 0.002 } } ] }

Request

Responses