API reference

A clean JSON API to manage streams and destinations, pull recordings, clips, analytics and highlights, and drive the unified cross-platform chat — programmatically.

Base URL

https://streamrepeater.com/api

curl https://streamrepeater.com/api/openapi.json -o streamrepeater-openapi.json

Authentication

All endpoints are authenticated using a Sanctum Personal Access Token. You can generate or revoke these tokens from your dashboard under Settings → API tokens. Requests are rate-limited to 60 / minute per IP, and all errors are returned as JSON.

Send the token in the Authorization header as a Bearer token:

curl -H "Authorization: Bearer YOUR_TOKEN_HERE" https://streamrepeater.com/api/channels

Your account is scoped to the workspaces you belong to. Tokens can be read-only or read/write. Mutating stream and destination endpoints require a write token and the Manage streams role in that workspace. Workspace and stream provisioning endpoints are admin-only and are documented in the admin console.

Errors & conventions

Errors return a JSON body with a message; validation failures (422) also include an errors map keyed by field. Other common statuses: 401 (missing/invalid token), 403 (token lacks the ability, role or workspace access), 404 (not found or not visible to you). Growing feeds — chat and events — accept ?after=<id> to poll only newer rows.

Endpoints

AI Streams

Create and run always-on channels made from AI-generated video, steer them while they are on air, and read why they are (or are not) generating. These endpoints need AI Streams on the workspace's plan; writes need a write token and the Manage streams role. Creating and editing use exactly the same checks as the dashboard form, and a new stream is never switched on until you call start. Provider API keys are added on the dashboard only — the API tells you which providers are connected and whether each key works, never the key. See the AI Streams help articles for how the product behaves.

Push a topic into a live stream. The headline use: something just happened in your world — a goal, a sale, a breaking story — and you want the channel to be about it. The topic takes over for minutes (default 30), then the stream returns to its running order exactly where it left off. It does not jump the queue: clips already made play first, and the reply says how many (clips_ahead). It never starts a stopped stream and never makes a stream spend faster — the stream still asks for at most its clips ahead at a time, and every topic passes the same content check and spend limit.

curl -X POST -H "Authorization: Bearer YOUR_TOKEN_HERE" -H "Content-Type: application/json" \
  -d '{"instruction":"Fans erupt as the home side scores","minutes":10,"label":"Goal! 1-0"}' \
  https://streamrepeater.com/api/v1/tenants/1/ai-streams/5/topic

{
  "ok": true,
  "entry": { "id": 21, "kind": "date", "start_time": "19:42", "timezone": "UTC", "duration_minutes": 10 },
  "on_air_now": true,
  "clips_ahead": 2,
  "stream_active": true,
  "message": "The next clip this stream commissions will be about this topic. 2 clips already queued or being made will play first."
}

Play a particular beat next with POST …/beats/{beat}/cue, or add one and cue it in one call with {"instruction": "…", "cue": true}. Why is it quiet? GET …/status returns the same reason the stream page shows — stopped, not-entitled, no-provider, no-prompt, no-key, spend-limit, or null when nothing blocks it — with the sentence to show and what to do.

# Create a stream (never starts it), then start it
curl -X POST -H "Authorization: Bearer YOUR_TOKEN_HERE" -H "Content-Type: application/json" \
  -d '{"name":"Harbour cam","channel_ids":[3],"provider":"luma","prompt":"A calm harbour at dawn","clip_seconds":8,"aspect":"16:9","buffer_target":3,"budget_limit":50,"clip_cost_estimate":0.4}' \
  https://streamrepeater.com/api/v1/tenants/1/ai-streams
curl -X POST -H "Authorization: Bearer YOUR_TOKEN_HERE" https://streamrepeater.com/api/v1/tenants/1/ai-streams/5/start

Destination parameters

POST and PUT destination requests accept either a custom destination or a saved preset payload. For a custom destination, send name, protocol (rtmp, rtsp or srt) and url. Optional video_settings can include resolution (source, 1080p, 720p, 480p) and bitrate in Kbps. The toggle endpoint accepts {"enabled": true} or {"enabled": false}.

Example: list streams

curl -H "Authorization: Bearer YOUR_TOKEN_HERE" https://streamrepeater.com/api/channels

[
  {
    "id": 1,
    "path_name": "acme--main",
    "status": "live",
    "tenant": { "id": 1, "name": "Acme" },
    "destinations": [
      { "id": 7, "name": "YouTube", "protocol": "rtmp",
        "relay_process": { "status": "running" } }
    ]
  }
]

Example: add a destination

curl -H "Authorization: Bearer YOUR_TOKEN_HERE" \
                  -H "Content-Type: application/json" \
                  -d '{"name":"Twitch","protocol":"rtmp","url":"rtmp://live.twitch.tv/app/KEY"}' \
                  https://streamrepeater.com/api/channels/1/destinations

Example: update a destination

curl -X PUT -H "Authorization: Bearer YOUR_TOKEN_HERE" \
                  -H "Content-Type: application/json" \
                  -d '{"name":"YouTube backup","protocol":"rtmp","url":"rtmps://a.rtmps.youtube.com:443/live2/KEY","video_settings":{"resolution":"720p","bitrate":3500}}' \
                  https://streamrepeater.com/api/channels/1/destinations/7

Example: pause a destination

curl -X PATCH -H "Authorization: Bearer YOUR_TOKEN_HERE" \
                  -H "Content-Type: application/json" \
                  -d '{"enabled":false}' \
                  https://streamrepeater.com/api/channels/1/destinations/7/toggle

Example: delete a destination

curl -X DELETE -H "Authorization: Bearer YOUR_TOKEN_HERE" \
                  https://streamrepeater.com/api/channels/1/destinations/7

{
  "message": "Destination pending deletion.",
  "destination": { "id": 7, "enabled": false, "pending_deletion": true }
}

Example: channel status

curl -H "Authorization: Bearer YOUR_TOKEN_HERE" https://streamrepeater.com/api/channels/1/status

{
  "channel_id": 1,
  "path_name": "acme--main",
  "status": "live",
  "viewers": 7,
  "destinations": [
    { "id": 7, "name": "YouTube", "enabled": true, "relay_status": "running" }
  ]
}

Example: delete a channel

curl -X DELETE -H "Authorization: Bearer YOUR_TOKEN_HERE" \
                  https://streamrepeater.com/api/channels/1

{
  "message": "Channel pending deletion.",
  "channel_id": 1,
  "pending_deletion": true
}

Example: channel analytics

curl -H "Authorization: Bearer YOUR_TOKEN_HERE" https://streamrepeater.com/api/channels/1/analytics

{
  "channel_id": 1,
  "window": "24h",
  "samples": [
    { "sampled_at": "2026-06-26T12:00:00Z", "inbound_bps": 5200000, "outbound_bps": 10400000, "viewers": 12 }
  ]
}

Example: list highlights

curl -H "Authorization: Bearer YOUR_TOKEN_HERE" https://streamrepeater.com/api/channels/1/highlights

[
  { "id": 4, "occurred_at": "2026-06-26T14:32:00Z", "messages": 47, "intensity": 60,
    "sample": [ { "author": "fan123", "message": "LETS GOOO" } ] }
]

Unified chat

Read the live cross-platform chat feed, reply as the broadcaster, and moderate — the same engine as the dashboard chat panel. Replies support twitch, kick and youtube (where connected with the needed scope); moderation actions are delete, timeout (10 min) and ban. Chat endpoints require the workspace to have integrations enabled, and writes require the Manage streams role.

# Poll new messages since the last id you saw
curl -H "Authorization: Bearer YOUR_TOKEN_HERE" \
  "https://streamrepeater.com/api/tenants/1/chat?after=1024"

# Reply as your channel
curl -H "Authorization: Bearer YOUR_TOKEN_HERE" -H "Content-Type: application/json" \
  -d '{"platform":"twitch","message":"thanks for the raid!"}' \
  https://streamrepeater.com/api/tenants/1/chat/reply

# Time out a message's author for 10 minutes
curl -H "Authorization: Bearer YOUR_TOKEN_HERE" -H "Content-Type: application/json" \
  -d '{"action":"timeout"}' \
  https://streamrepeater.com/api/tenants/1/chat/messages/2048/moderate

Integrations & automations

Read your connected platforms, automation rules and event history — and, most usefully, fire your own events into the rules engine. A custom event runs every matching rule's actions (lights, Discord, webhooks, OBS…), so an external system — a donation page, your backend, a Zap — can drive your on-stream automations. Secrets are never returned: connection tokens and action payloads are omitted from the responses.

# Fire a custom event — runs every rule listening for "custom.donation"
curl -H "Authorization: Bearer YOUR_TOKEN_HERE" -H "Content-Type: application/json" \
  -d '{"type":"custom.donation","context":{"name":"Ada","amount":"5.00"}}' \
  https://streamrepeater.com/api/tenants/1/events

{ "ok": true, "matched_rules": 2 }

# Read the recent event feed
curl -H "Authorization: Bearer YOUR_TOKEN_HERE" "https://streamrepeater.com/api/tenants/1/events?after=900"

Create rules that listen for custom.* events in Integrations → Automations, then reference your context fields (e.g. {{ amount }}) in the action templates.

Event webhooks

Create developer webhooks from Settings -> Developer webhooks. Each delivery is a JSON POST signed with X-Streamrepeater-Signature, where the value is sha256= plus an HMAC-SHA256 of the raw request body using the endpoint's signing secret.

{
  "id": "9b6f6b4d-2d13-4d4b-b3b8-54ff5d29c9ef",
  "event": "stream.live",
  "created_at": "2026-06-16T12:00:00+00:00",
  "data": {
    "channel": { "id": 1, "title": "Main Feed" },
    "destination_count": 2
  }
}

Available events: stream.live, stream.offline, destination.failed, recording.started and bandwidth.warning; and for AI Streams ai_stream.started, ai_stream.stopped, ai_stream.yielded (paused for a live broadcast), ai_stream.channel_unavailable, ai_stream.video_saved (at most hourly per stream), ai_provider.key_refused and ai_provider.out_of_credit — each carries a message, details and the stream or provider it is about, and repeats are collapsed the way the notification bell collapses them. The same AI events can trigger automation rules. The settings page can also send a ping test event. Endpoints can be created, listed and deleted via the API (/tenants/{tenant}/webhooks) as well as the dashboard — the signing secret is returned once on creation.

Need help integrating? Get in touch.