Skip to main content
These limits apply to the hosted Agent API, including requests made through the CLI and MCP server. Credits, request limits, and concurrent jobs are separate controls.

​
Request limits

Each bucket is checked both per API key and per owning account. Multiple keys do not multiply the account allowance. Windows are fixed 60-second periods; unused capacity does not carry forward. A connection to an event stream counts as one request; individual events are not separate HTTP requests. Other application surfaces can have their own limits. This is not a throughput or burst guarantee.

​
Concurrent work

  • Up to two active managed image jobs per account, including sprite-sheet jobs.
  • Up to two active Video admissions per account, additionally subject to shared service capacity. This is a separate Video allowance, not two extra image slots.
  • Up to four open image/sprite execution event streams per API key, or three Video streams per API key. Close disconnected or unused streams before opening more.
  • A closed stream does not cancel a job or release its credit hold. Check execution state and reconnect to the same execution.
Keep your own queue. Start with one active job, especially when another application shares the account. Two simultaneous submissions can still exceed a request window when uploads, polling, and other callers share the same account.

​
Response headers

Errors include X-Request-ID, Cache-Control: no-store, and Pragma: no-cache. Keep the request ID for support. The current public contract does not promise Retry-After, remaining-request headers, or reset-time headers. Do not depend on their presence. If a future response includes Retry-After, honor it; otherwise use capped exponential backoff with jitter. A practical client policy starts at one second, doubles between attempts, caps at 60 seconds, and stops after a bounded number of retries. That is client guidance, not a server completion deadline.

​
Decide what to do

For image and sprite requests to /v1/execute, distinguish synchronous HTTP errors from later execution failures. A 202 or open SSE stream does not guarantee admission. Non-retryable admission errors are returned in canonical execution state and done; credit and quota failures do not return submit-time 402 for those requests. Video is different: accepting a quote can immediately return 402 insufficient_credits or 429 too_many_active_jobs. A successful Video acceptance is 202 JSON. See Video errors and recovery. The table below describes image/sprite admission. Keep the same uploaded input_asset_id on a replay. Uploading the file again creates a different request, even when the local file is unchanged. A terminal failure is a completed attempt: submitting a replacement is a new operation that can incur a new charge. See idempotency and retries, error details, and the multiple-image queue.

​
Prompt lengths

Image prompts, Video quote prompts, and sprite options.animation_prompt each have a 4,000-character limit, including spaces and line breaks. Sprite animation instructions must be nonblank and cannot be combined with a preset action. See sprite options and Video settings.