Rate Limits - Framelane

Limits

Endpoint	Limit
`POST /v1/signup`	10 requests / IP / hour
`POST /v1/signup/verify`	30 requests / IP / hour
`POST /v1/auth/sync`	10 requests / IP / hour
`POST /v1/renders`	120 requests / IP / minute
`POST /v1/tasks/*`	300 requests / IP / minute

Only the last three are machine-readable — GET /v1/capabilities exposes them under rate_limits as the auth_sync, render_submit, and task_submit scopes. The two signup limits are documented here only. Workspace-level monthly quotas (applies to the free plan):

Resource	Free plan
Billable render time	30 min / month

Previews (POST /v1/preview, and any is_preview job) are never metered, and only completed renders count toward the cap. There is no quota on AI tasks.

Rate limit headers

When a limit is exceeded, the API returns 429 Too Many Requests:

{
  "error": {
    "code": "rate_limited",
    "message": "Too many requests. Retry after 47 seconds."
  }
}

Quota exceeded

When the free-plan render-time cap is exhausted, POST /v1/renders returns 402 Payment Required with code: quota_exceeded — not 429. It is a monthly cap, not a rate limit: do not retry it with backoff.

{
  "error": {
    "code": "quota_exceeded",
    "message": "Free-tier render limit reached (30 min/month). Used 29.4 min this period; this render is 1.2 min. Upgrade your plan to render more.",
    "details": {
      "limit_seconds": 1800,
      "used_seconds": 1764.0,
      "requested_seconds": 72.0
    }
  }
}

Upgrade your plan in the console to increase limits.

Best practices

Retry 429 with exponential backoff; never retry 402 quota_exceeded that way — it clears on the monthly reset, not on a delay.
Use idempotency keys to safely retry without double-submitting.
For agent workflows, use the batch-tasks Agent Skill to submit multiple tasks concurrently without hitting per-request limits.
Subscribe to webhooks instead of polling to avoid unnecessary GET requests.