Errors | Kontinent | Documentation

Errors are always OpenAI-shaped:

1 { "error": { "message": "…", "type": "…", "code": "…" } }

Status reference

HTTP	`code`	`type`	Meaning / client action
400	`invalid_request`	`invalid_request_error`	Malformed body, over 2 MiB, or missing `model`.
401	`invalid_api_key`	`invalid_request_error`	Bad or revoked key — do not retry.
402	`insufficient_credits`	`insufficient_quota`	Balance ≤ 0 — top up, then retry.
404	`model_not_found`	`invalid_request_error`	Unknown or disabled model id.
429	`rate_limit_exceeded`	`rate_limit_error`	Per-key RPM hit — back off (sliding 60 s window).
429	`upstream_rate_limited`	`rate_limit_error`	Provider throttled us — retry with backoff.
500	`internal_error`	`api_error`	Gateway bug — retriable.
502	`upstream_auth_error`	`api_error`	Provider rejected our credentials — our problem.
502	`upstream_error`	`api_error`	Provider 5xx / connect failure — retriable.

Passthrough errors

Upstream 4xx bodies that are already OpenAI-shaped (for example, provider-side validation of your parameters) are passed through with their original status rather than rewrapped.

Retry guidance

Retriable

429, 500, 502 upstream_error. Use exponential backoff with jitter.

Not retriable as-is

400, 401, 404. Fix the request or the key first. 402 is retriable only after a top-up.