API Overview

Base URL

All API requests are made to:

https://api.getsieve.dev

Authentication

Authenticate by including your API key in the request header. Sieve uses prefixed API keys:

• mod_live_ — Production keys. Calls are billed and routed through the full moderation pipeline.
• mod_test_ — Test keys. Returns deterministic responses, makes no AI calls, and incurs no charges.

Authorization: Bearer mod_live_abc123...

See Authentication for full details.

Rate Limits

Rate limits are enforced per API key and vary by plan:

Plan	Rate Limit	Monthly Calls
Free	10 req/sec	10,000
Indie ($29/mo)	50 req/sec	100,000
Pro ($99/mo)	200 req/sec	1,000,000

When you exceed the rate limit, the API returns a 429 Too Many Requests response with a Retry-After header.

Test Mode

Any request made with a mod_test_ key enters test mode:

• Returns deterministic responses based on known phrases and patterns
• Makes no downstream AI calls (OpenAI, Claude)
• Incurs no billing — test calls are not counted toward your quota
• Useful for integration testing, CI/CD pipelines, and local development

Response Format

All successful responses return JSON with a consistent structure:

{
  "request_id": "req_abc123",
  "action": "allow",
  "flagged": false,
  "categories": {
    "toxicity": {
      "score": 0.02,
      "threshold": 0.7,
      "flagged": false
    },
    "harassment": {
      "score": 0.01,
      "threshold": 0.7,
      "flagged": false
    }
  },
  "pipeline_tier": 0,
  "latency_ms": 3
}

Field	Type	Description
request_id	string	Unique identifier for the request, useful for support and debugging.
action	string	Recommended action: allow, flag, or block.
flagged	boolean	true if any category exceeded its threshold.
categories	object	Per-category scores, thresholds, and flags.
pipeline_tier	integer	Which pipeline tier handled the request (0 = local, 1 = OpenAI, 2 = Claude).
latency_ms	integer	Total processing time in milliseconds.

Errors

The API uses standard HTTP status codes. Error responses include a JSON body:

{
  "error": {
    "code": "rate_limit_exceeded",
    "message": "Rate limit exceeded. Retry after 1 second.",
    "retry_after": 1
  }
}

Status	Meaning
400	Bad request — invalid or missing parameters.
401	Unauthorized — missing or invalid API key.
429	Rate limit exceeded.
500	Internal server error.