Fix a 429 error (too many requests)

A 429 means “slow down”: the request was turned away because of how fast requests are arriving, not because of your plan or balance. Wait the number of seconds in the Retry-After header, then try again.

The three 429 codes

Code	Cause	Fix
`rate_limited`	More than 600 requests in a minute on one key (or one Syntic Code sign-in)	Wait `Retry-After` seconds; smooth out bursts
`model_busy`	The model is handling too many requests right now	Retry after `Retry-After`; for non-urgent work, try another model
`capacity_busy`	Syntic turned the request away before running it. Nothing was charged	Retry after `Retry-After`

The error body also includes retry_after (in seconds), the same value as the Retry-After header.

Retry the right way

Read Retry-After and wait at least that long.
Add a little random delay (jitter) so many clients don’t all retry at the same moment.
If the next try also gets a 429, wait longer each time (exponential back-off), up to a sensible cap.
Give up after a few attempts and log the request ID.

Stay under the limit

Send requests through a queue in your code instead of all at once.
Use one key per service so a busy service doesn’t use up another’s allowance. Note that extra keys aren’t a way around the limit for the same workload.
Cache answers to repeated questions.

Need a higher limit?

In the Console, choose Contact Sales, pick Increase Rate Limits, and describe your expected traffic (requests per minute at peak, and which models).

Not a 429?

402 is about plan windows and credit: Fix a 402 error.
503 errors such as service_unavailable mean a temporary problem; retry with back-off and check status.syntic.ai.

Fix a 402 error Common error codes

Was this page helpful?