For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
Logo
DocumentationAPI ReferenceChangelog
DocumentationAPI ReferenceChangelog
  • Get Started
    • Introduction
    • Authentication
    • Versioning
    • SDKs
    • x402
    • Compute Units (CU)
    • Prompt Guidance
    • Errors
    • Rate Limits
    • Pagination
On this page
  • Exceeding limits
Get Started

Rate Limits

Built with

If you require higher limits, please contact us at support@caesar.xyz.

Scope: These limits apply only to POST /research (create research). Other endpoints are not rate-limited under this policy.

The Caesar API has the following limits:

  • Concurrent research jobs: 5 per account (in-progress jobs, including queued).
  • Monthly compute budget: 200 compute units, counted only for successful research jobs.

Exceeding limits

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

1{ "error": "Too many requests - rate limit exceeded." }

Recommended client behaviour

  • Retry with exponential back-off.
  • Serialise new jobs when you’re at the concurrency cap.
  • Use compute_units thoughtfully to stay within your monthly budget.
What is a compute unit?

A compute unit represents the effort allocated to a research job. Your monthly usage is the sum of the compute_units for jobs that reach the completed status.

What counts as 'concurrent'?

Any job that is not yet completed or failed (e.g. queued, searching, summarising, analysing) counts towards your concurrency limit.

When do monthly limits reset?

Limits reset on a monthly cadence; your dashboard shows your remaining balance and the next reset window.

Can I request higher limits?

Yes. Email support@caesar.xyz with your expected volumes and use case.

Where can I see my remaining quota?

Visit your dashboard after signing in.