Get a Zoom transcript via API

Send a bot to the meeting with one POST, then read the transcript while the call is still running. The bot joins Zoom like any participant — no plugins, no host configuration — and streams back a speaker-attributed transcript you either poll over REST or subscribe to over WebSocket.

Hosted, no install: connect your agent (Claude Code, Codex, Cursor) or get an API key (free credit, no card required). Self-hosting: Deployment.

1. Send the bot

curl -X POST "$API_BASE/bots" \
  -H "X-API-Key: $API_KEY" -H "Content-Type: application/json" \
  -d '{"platform":"zoom","native_meeting_id":"81234567890","bot_name":"Vexa"}'

native_meeting_id is the meeting id inside the join URL, not the whole URL. Have only the URL? Send meeting_url alone instead of platform + native_meeting_id — the API parses it to extract both, and, for a Zoom link that embeds one, the passcode. A URL that is not a recognizable meeting link is refused with a 422 naming the missing native_meeting_id. $API_BASE is https://api.cloud.vexa.ai hosted, or http://localhost:18056 self-hosted. Optional on the spawn: language (ISO code; omit for per-window auto-detection) and task (transcribe default, or translate).

2. Read the transcript live

Segments stream in as the meeting runs. Poll:

curl -H "X-API-Key: $API_KEY" \
  "$API_BASE/transcripts/zoom/81234567890"

{
  "segments": [
    { "speaker": "Jane Liu", "text": "Let's lock the renewal pricing by July 1.",
      "start": 12.4, "end": 15.1, "language": "en", "completed": true, "confidence": 0.93 }
  ]
}

Live drafts arrive as completed: false and are replaced by completed: true confirmations. For per-segment push instead of polling, subscribe on the gateway’s /ws:

{ "action": "subscribe",
  "meetings": [ { "platform": "zoom", "native_id": "81234567890" } ] }

Zoom specifics

Web client only. In 0.12.x the bot joins Zoom through the web client. The native SDK / OBF-token path is not carried — see Roadmap → Status.
Passcodes. The field is spelled passcode. password, meeting_password and the other near-misses are refused with a 422 naming the field to use, rather than accepted and dropped. A passcode already in the URL’s query wins over a separate passcode field.
Language auto-detection. Zoom rides the mixed-audio lane, where a low-confidence detection discards the window rather than guessing; those windows carry language: null. Pin language on the spawn to avoid it.
Leaving on its own. An active bot leaves after 10 minutes with no remote microphone audio and finishes as completed(left_alone). Widen it per meeting with automatic_leave.max_time_left_alone.

Stop the bot

curl -X DELETE "$API_BASE/bots/zoom/81234567890" -H "X-API-Key: $API_KEY"

Start here

How-to: Meetings

How-to: Agents

How-to: Meetings + agents

Core

Concepts

Architecture

API reference

Deployment & operations

Roadmap

Governance

Get a Zoom transcript via API

1. Send the bot

2. Read the transcript live

Zoom specifics

Stop the bot

See also

​1. Send the bot

​2. Read the transcript live

​Zoom specifics

​Stop the bot

​See also

1. Send the bot

2. Read the transcript live

Zoom specifics

Stop the bot

See also