Per-Participant Video Streams: Separate Video for Every Participant

Get one video file per participant webcam and per screen share, instead of (or as well as) a single composite recording of the meeting.

Supported platforms: Zoom, Google Meet, Microsoft Teams.

Platform support

When more cameras are on than the platform limit, the extra cameras are not recorded until a slot frees up. Plan around the limit on Teams in particular: in a meeting with more than six cameras on, some participants will not have a video file.

See the capability matrix for the rest of what each platform supports.

Per-participant video is opt-in and off by default. A bot never records one file per participant unless you set video_separate_streams: true yourself, because it multiplies processing and storage by the number of participants. Leave it unset for ordinary recording.

1) What you need

1. A MeetStream API key (see Create your first bot).
2. A meeting link (Zoom, Google Meet or Microsoft Teams).
3. video_separate_streams: true in your Create Bot request. On Zoom, also video_required: true.

2) Create a bot with per-participant video

Use the Create Bot endpoint.

Minimal example

curl -X POST "https://api.meetstream.ai/api/v1/bots/create_bot" \
  -H "Authorization: Token <YOUR_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "meeting_link": "<YOUR_MEETING_LINK>",
    "bot_name": "RecorderBot",
    "video_required": true,
    "video_separate_streams": true
  }'

Sending video_required: true works on every platform and is required on Zoom.

To try it without code, open API Playground in the dashboard and go to the Streams step:

Full example with transcription

{
  "meeting_link": "https://meet.google.com/abc-defg-hij",
  "bot_name": "RecorderBot",
  "video_required": true,
  "video_separate_streams": true,
  "recording_config": {
    "video_layout": "speaker_view",
    "transcript": {
      "provider": {
        "deepgram": {
          "model": "nova-3",
          "language": "en"
        }
      }
    }
  },
  "automatic_leave": {
    "everyone_left_timeout": 60
  }
}

The composite recording here uses video_layout: "speaker_view" so the single MP4 follows whoever is talking, which is what you want for review and clipping. The per-participant WebM files are unaffected by the layout.

Parameter reference

Tip: You can also pass video_separate_streams inside recording_config:

"recording_config": {
  "video_separate_streams": true
}

The top-level parameter takes precedence if both are set.

3) What happens during the meeting

Once the bot is admitted, it automatically:

• Captures each participant’s webcam as a separate stream.
• Captures each screen share as a separate stream.
• Keeps each file on the meeting’s timeline, so files line up with the mixed audio and with each other.

No per-participant configuration is needed. If you pause recording, every per-participant file shows black video for the paused stretch.

4) Retrieve per-participant video streams

After the bot leaves and video processing finishes (the video.processed webhook), call:

curl -X GET "https://api.meetstream.ai/api/v1/bots/<BOT_ID>/get_recording_streams" \
  -H "Authorization: Token <YOUR_API_KEY>"

Response when processing is complete (200)

The response is an array. Each item is one recorded webcam or screen-share stream:

[
  {
    "id": "f923cd21-da86-4d77-b37b-0707d37751c8-user_456-0",
    "created_at": "2026-09-20T17:02:11.482Z",
    "participant": {
      "id": 48213377,
      "name": "Jane Smith",
      "is_host": true,
      "platform": "desktop",
      "extra_data": {
        "google_meet": { "static_participant_id": "spaces/AbCdEf/devices/185" }
      }
    },
    "start_timestamp": { "absolute": "2026-09-20T17:02:11.482Z", "relative": 12.4 },
    "duration": 1325.5,
    "download_url": "https://...",
    "type": "webcam"
  },
  {
    "id": "f923cd21-da86-4d77-b37b-0707d37751c8-source_101-0",
    "created_at": "2026-09-20T17:10:40.006Z",
    "participant": {
      "id": 48213377,
      "name": "Jane Smith",
      "is_host": true,
      "platform": "desktop",
      "extra_data": {
        "google_meet": { "static_participant_id": "spaces/AbCdEf/devices/185" }
      }
    },
    "start_timestamp": { "absolute": "2026-09-20T17:10:40.006Z", "relative": 520.9 },
    "duration": 60.0,
    "download_url": "https://...",
    "type": "screenshare"
  }
]

Other responses

5) Download the video files

curl -o "jane-webcam.mp4" "<download_url from the response>"

Links expire after 1 hour. Call get_recording_streams again for fresh links.

Audio is separate: use per-participant audio or the mixed recording from retrieve recordings.

6) Understanding the files

• A participant can have more than one item, for example a webcam file and a screen-share file, or one file per screen-share session.
• Turning a camera off and on does not start a new file: the off stretch is filled with a black tile showing the participant’s name, so the file stays on the meeting timeline.
• Use start_timestamp.relative and duration to place each file on the meeting timeline, and participant.extra_data to match it with per-participant audio.

7) Using both composite and per-participant video

{
  "meeting_link": "<MEETING_LINK>",
  "video_required": true,
  "video_separate_streams": true,
  "recording_config": {
    "video_layout": "speaker_view"
  }
}

video_layout applies only to the composite MP4. Speaker view is the recommended setting because it follows the active speaker; switch to grid_view if you specifically want a mosaic of everyone.

This produces:

• A composite MP4 of the meeting (layout grid_view or speaker_view, built after the call from the per-participant streams). Retrieve it with Get Bot Video.
• Per-participant MP4 files. Retrieve them with Get Recording Streams.

Pick the composite layout with recording_config.video_layout, or on the Playground’s Record step:

8) Webhook notifications

Subscribe to video.processed (and video.failed) in webhooks and events, then call get_recording_streams. Polling until you get a 200 with items works too.

9) Troubleshooting

• Empty array [] after the meeting?

  • Check that video_separate_streams: true was set, and on Zoom that video_required: true was set as well.
  • Participants who never turned their camera on and never shared their screen do not produce files.

• A participant’s video is missing?

  • The platform limit was reached: 6 streams on Teams (including 1 screen share), 16 on Zoom and Google Meet. Extra cameras are not recorded until a slot frees up.
  • The bot only records after it is admitted. Anything before admission is not captured. See bot status and error codes if the bot was never admitted.

• Getting 202 with video_status: "processing"?

  • Processing takes a short time after the bot leaves. Poll again after 10 to 30 seconds.

• The download link returns 403 Forbidden?

  • The link has expired (1-hour lifetime). Call get_recording_streams again.

• The video file has no audio?

  • Per-participant video files are video-only. See per-participant audio.