REST API

Route descriptions

The module lets you retrieve completed transcripts and perform other actions through its REST API. This can be useful for integrations with external systems. The most important REST API information is provided below. Base path:

/pbxcore/api/v3/module-cloud-speech-to-text

Purpose	Endpoint
Summary and status	GET /dashboard, GET /health
Selection rules	GET/PATCH /rule-sets/current
Mode and retention	GET/PATCH /settings
Connection check	POST /connection-checks
Processing control	GET/PATCH /processing-controls
Job queue	GET /jobs
Job cancellation and retry	POST /job-cancellations
Transcript list	GET /transcripts
Transcript export	GET /transcript-exports/{id}
Module log	GET /logs
Diagnostic archives	GET/POST /diagnostic-reports
Transcript lookup by call	GET /cdr-transcript-lookups

Integrations: call-transcripts

External modules read completed transcripts through two dedicated resources: GET /call-transcripts and GET /call-transcript-events. These resources accept only a dedicated integration key and return the complete immutable transcript contract: segments, merged turns, and plain text.

The identification scheme ensures that an external system does not repeat its analysis after a clean reinstallation of the module:

• call_id — a stable Asterisk call identifier (linkedid, or UNIQUEID when linkedid is empty);

• source_instance_id — the module installation identifier;

• semantic_hash — a hash of the transcript content using the call-transcript-analysis-v1 profile: it matches only if the text, participants, and timings have not changed.

The call-transcript.completed, call-transcript.updated, and call-transcript.deleted events are delivered through cursor-based pagination and are idempotent. The public contract schema is call-transcript.v1. After a transcript is deleted, events no longer expose call_id or hashes.