StreamLive Pro

API rate limits, retries and Idempotency-Key

StreamLive Pro API rate limits per key, per account, for writes and for publishing; the RateLimit headers, what to do on 429, and safe retries with Idempotency-Key.

Updated November 5, 2026 · 4 min read

The StreamLive Pro API limits how fast one key and one account can call it, so a loop in someone's code cannot flood your channels. The limits are 120 calls a minute per key or connection, 240 a minute per account across all keys, 30 writes a minute per account, and 20 publishes an hour per account. Every answer tells you where you stand in headers, and a refused call says how many seconds to wait. For writes, send an Idempotency-Key header so a retry never does the same thing twice.

Keys, permissions and the list of endpoints are in API keys for developers: create, use and revoke. The same operations as assistant tools: StreamLive Pro MCP tools: full list for ChatGPT and Claude.

The limits

The windows are fixed: a minute window resets at the start of the next minute, the publishing window at the start of the next hour. The same limits apply to ChatGPT and Claude, since they run the same operations. Your plan's limits, and the networks' own rules against spam, apply on top.

Headers on every answer

When you are over a limit, the answer is 429 with type rate_limit_error and code rate_limited, and a message such as "Rate limit reached. Try again in 12 seconds."

Steps

1. Read RateLimit-Remaining. When it gets low, slow down before you hit zero.
2. On 429, wait. Sleep for the seconds in Retry-After, then try again. Do not retry at once in a loop.
3. Send an Idempotency-Key on every write. A unique value per action, 8 to 100 characters of letters, digits, underscore, dot, colon or hyphen. For example publish-clip_123-2026-11-05.
4. Retry server errors with the same key. On a timeout, 500, 502 or 503, wait a little, then send the exact same request with the same Idempotency-Key. Double the wait each time.
5. Do not retry most 4xx answers as they are. A 400, 401, 403 or 404 will fail again until you fix the request. A 409 means the state changed: read the object again, then decide.
6. Keep the Request-Id. Log it with each error. Support can find the request by it.

How Idempotency-Key works

• It applies to POST, PATCH and DELETE.
• Once a request with a key succeeded, for 24 hours the same key with the same request returns that first answer again, with the same status, instead of doing the action again. The answer carries the header Idempotent-Replayed: true. A request that failed is not remembered, so you can fix the cause and retry with the same key.
• The same key with a different request (another path or other fields) is refused with 422 idempotency_key_reused.
• A key in the wrong format is refused with 400 invalid_idempotency_key.
• Keys are per account, so two accounts can use the same value.

This matters most for POST /posts, which publishes at once with no preview, and for POST /lives. Without the header, a retry after a timeout can publish the same Short twice.

curl -X POST https://streamlivepro.io/api/v1/posts \
  -H "Authorization: Bearer slp_..." -H "Content-Type: application/json" \
  -H "Idempotency-Key: publish-clip_7f3a-2026-11-05" \
  -d '{"clip_id":"clip_...","destination_ids":["dest_..."],"title":"Five tips","youtube_privacy":"public"}'

What can go wrong

The error format and the first codes you will meet are in API keys for developers: create, use and revoke. Plan limits that apply on top: When you reach a limit: what the message means.

Frequently asked questions

What is the rate limit of the StreamLive Pro API?

120 calls a minute per key, 240 a minute per account, 30 writes a minute per account and 20 publishes an hour per account.

What should my code do on a 429?

Wait the number of seconds in the Retry-After header, then send the request again.

How long is an Idempotency-Key remembered?

24 hours after a successful request. The same key with the same request returns the first answer instead of repeating the action.

Do ChatGPT and Claude count toward the same limits?

Each assistant connection has its own per-connection limit, and all of them count toward the account limits together with your API keys.

Can I ask for higher limits?

The limits are the same for every account today. Write to [email protected] with what you are building if they block a real use.

Keep reading