Skip to main content

Documentation Index

Fetch the complete documentation index at: https://www.rhetoricaudit.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

The Rhetoric Audit API lets you programmatically submit article text for Forensic Media Evaluation (FME) and retrieve the resulting analysis. All endpoints are served over HTTPS, respond with JSON, and are CORS-open so you can call them from any origin including browser extensions and client-side JavaScript.

​
Base URL

https://www.rhetoricaudit.com

​
Available endpoints

MethodPathDescription
POST/api/analyzeSubmit text for FME analysis
GET/api/analyzeHealth check — returns API version and status
GET/api/scans/{shortcode}Retrieve a cached scan by shortcode
POST/api/scans/{shortcode}/privacyToggle the public/private flag on a scan
POST/api/intelligence/analyzeGenerate an Intelligence Brief (multi-source narrative analysis)
GET/api/intelligence/briefsList your historical Intelligence Briefs

​
Response format

Every endpoint returns Content-Type: application/json. Error responses always include an error string field. Successful responses include the fields described on each endpoint’s page.

​
Versioning

The current FME engine version is V19.1 and the API is at v3.0.0. You can confirm both values at any time with the health check endpoint. The fmeVersion field is present in every analysis response.

​
CORS

All endpoints return Access-Control-Allow-Origin: *. You do not need a proxy to call the API from a browser or extension.

​
Rate limits

Rate limits apply only to cache-miss requests — requests that return a cached result are always free and never counted against any quota.
TierIdentityDaily limit
anondeviceId only10 requests/day
signed_freeuserId, 0 credits50 requests/day
paiduserId, credits > 0Unlimited (1 credit per cache miss)
When a rate limit is exceeded the API responds with HTTP 429 and a JSON body that includes rateLimited: true, tier, limit, used, and resetAt (ISO 8601 timestamp of the next daily reset at UTC midnight).

​
Health check

curl https://www.rhetoricaudit.com/api/analyze
Response 
{
  "status": "ok",
  "version": "3.0.0",
  "fme_prompt_version": "V18.6",
  "primary_model": "openai/gpt-4o-mini",
  "fallback_model": "openai/gpt-4.1-nano",
  "cache": "enabled",
  "service": "Rhetoric Audit FME Analysis API",
  "timestamp": "2026-05-09T12:00:00.000Z"
}
The fme_prompt_version field in the health check response reflects the active prompt version. The fmeVersion field in individual analysis responses may show a more specific version string that includes the pipeline variant used for that particular scan.