Documentation

Errors

What each status means, and what to change.

Common errors

StatusErrorMeaning
401 missing_provider_key The X-LLM-API-Key header was not sent.
402 free_tier_exhausted Your account used up its free tokens — 1M, or 2M once a card is on file. The response carries used, limit and an upgrade_url. See billing.
402 spend_limit_reached You set a monthly ceiling on our fee and this month has reached it. The response carries spent_usd and limit_usd. Nothing is owed and nothing is suspended — raise or clear the limit in the dashboard, or wait for the month to roll over.
402 billing_suspended An invoice went unpaid for more than 3 days. Pay it in the dashboard and access restores automatically.
401 / 403 (from your provider) Your provider key itself was rejected upstream — check it directly with OpenAI/Anthropic first.
400 / 404 (from your provider) The model name does not exist on your provider, or your key has no access to it. We forward the name as-is, so you get the provider's own error.
400 LLM Provider NOT provided The model name isn't recognisable as belonging to a known provider (typo, or a self-hosted model). Prefix it with the API flavour — openai/your-model — and send X-LLM-API-Base if it runs on your own server.