POST /lives

Creates a new live stream and returns the live object with the ingest endpoint for your encoder (OBS, etc.), the playback URLs and the IDs needed to broadcast.

📘

Authorization

Requires a valid Panda API token sent in the Authorization header (no Bearer prefix).

🚧

Maximum quality is 720p. Values above it — fhd, 4k, and the 1080p / 2160p entries of bitrate — are accepted and silently reduced to 720p. The live is created without an error. Set your encoder to 720p.

HTTP Method & Path

POST /lives

Base URL: https://api-v2.pandavideo.com.br

Authentication Requirements

Parameters

Request Body

Content-Type: application/json

Body Parameters

📘

The body rejects unknown fields

Any key outside the list above returns 400 naming the offending field.

🚧

No timezone suffix. Send 2026-09-20T14:30:00. A trailing Z (2026-09-20T14:30:00Z), which toISOString() always adds, is rejected. Scheduling is limited to 30 days ahead.

⚠️

One live per stream key. To run two lives at the same time you need two stream keys. How many keys you can create comes from your plan, so the limit shows up when creating a key — POST /live_stream_key returns 400 the plan has no slot available to create stream key — not when creating a live.

📘

region is where you publish, not where your viewers are. It selects the server that receives the signal from your encoder. Picking the wrong one only makes your encoder publish further away; delivery to viewers does not depend on this field. It also affects the per-minute rate, and if the stream key was created with a fixed region, the key's region wins.

⚠️

save_vod: false means there will be no video afterwards. The live runs normally and simply leaves no recording, and it cannot be recovered later.

{
  "title": "Aula ao vivo 01",
  "region": "us",
  "quality_package": "hd",
  "save_vod": true,
  "active_dvr": true,
  "latency_type": "default"
}

Response

Success Response (200)

{
  "id": "{live_id}",
  "title": "Aula ao vivo 01",
  "status": "provisioning",
  "stream_key_id": "{stream_key_id}",
  "stream_key": "{stream_key}",
  "user_id": "USER_ID",
  "region": "us",
  "quality_package": "hd",
  "save_vod": true,
  "bitrate": ["360p", "480p", "720p"],
  "active_dvr": true,
  "latency_type": "default",
  "ingest_endpoint": "rtmp://vz-{hash}.us.pandavideo.live/live/{stream_key}",
  "ingest_hostname": "vz-{hash}.us.pandavideo.live",
  "rtmp": "rtmp://vz-{hash}.us.pandavideo.live/live",
  "playback_url": null,
  "live_hls": "https://b-vz-{hash}.tv.pandavideo.com.br/{live_id}/playlist.m3u8.live",
  "live_player": "https://player-vz-{hash}.tv.pandavideo.com.br/embed/live/?v={live_id}",
  "scheduled_at": null,
  "started_at": null,
  "ended_at": null,
  "vod_id": null,
  "video_id": null,
  "folder_id": null,
  "view_metrics": {},
  "config": {},
  "chat_id": null,
  "chat_status": null,
  "created_at": "2026-09-07T02:37:22.591Z",
  "current_time": "2026-09-07T02:41:51.204Z"
}

📘

Two things worth noticing in the response

The live is created as provisioning, and playback_url comes back null at this moment — it is filled in once the live is ready. The ingest_endpoint, on the other hand, comes back right away, and it is what you need to configure your encoder.

Response Schema

Live status values

📘

expired is not a failure. If the live becomes ready but no signal ever arrives, it moves to expired. Do not create a new live: call POST /lives/{live_id}/regenerate and the same live goes back to provisioning, keeping its id, its stream key and its player URL.

Error Responses

400 Bad Request — the message names what went wrong:

// unknown field in the body
{ "errCode": "BadRequest", "errMsg": "must NOT have additional properties: X" }

// title shorter than 3 characters
{ "errCode": "BadRequest", "errMsg": "must NOT have fewer than 3 characters" }

// value outside the allowed list (the error lists what is accepted)
{ "errCode": "BadRequest", "errMsg": "must be equal to one of the allowed values: us,br." }

// scheduled_at with a timezone suffix, or more than 30 days ahead
{ "errCode": "BadRequest", "errMsg": "invalid scheduled_at property" }

// no valid stream key on the account
{ "errCode": "BadRequest", "errMsg": "Could not find any valid stream key" }

// trial account
{ "errCode": "BadRequest", "errMsg": "live is not available for trial accounts" }

// storage full
{ "errCode": "BadRequest", "errMsg": "storage limit reached. Free up space before creating a new live." }

⚠️

Three of these errors are returned in Portuguese. The API mixes languages, and this is the text your integration will actually receive:

• stream_key já está em uso por uma live ativa (online/provisionando) ou expirada — regenere a live expirada
  The key is in use by an active or expired live — regenerate the expired one.
• stream_key já está em uso por uma live ativa (online/provisionando)
  The key is in use by a live that is online or provisioning.
• stream_key já possui uma live aguardando transmissão
  The key already has a live waiting for the broadcast.

All three say the same thing by different routes: that key is already taken. Use another key, finish the live that is using it, or — if the live is expired — call the regenerate endpoint. This is the most likely error family in day-to-day use, because it is one live per stream key.

• 401 Unauthorized — missing or invalid API token.
• 500 Internal Server Error — server-side error.

Example Usage

cURL

curl -X POST "https://api-v2.pandavideo.com.br/lives" \
  -H "Authorization: {api_key}" \
  -H "Content-Type: application/json" \
  -d '{
    "title": "Aula ao vivo 01",
    "region": "us",
    "quality_package": "hd",
    "save_vod": true,
    "active_dvr": true,
    "latency_type": "default"
  }'