Searching for basics
Simple questions should have simple, findable answers.
Help center
Answers, right when you need them
Quick answers to the questions we hear most about API keys, errors, rate limits, caching, the free allowance and invoices. Still stuck? Our team is one email away.
Quick answers
to common questions
Self-serve
any time of day
Real people
when you need more
The problem
What is slowing you down today
Cryptic errors
An error message should tell you what to do next.
Billing confusion
You should always understand what you are paying for.
Capabilities
Everything you need, nothing you don't
API keys
Create keys in Settings. Each is shown once and stored as a hash, so a lost key cannot be recovered: create a new one and revoke the old.
401 Not signed in
Check the Authorization: Bearer <key> header and that the key has not been revoked.
429 Rate limit reached
Wait the seconds given in the error message, then retry. You can adjust your limit in Settings.
409 errors
The named model is disabled, or no enabled model offers the requested capability. Check the catalog.
Free allowance
Your free monthly tokens are shown on the Usage page. Metered pricing applies once they are used.
Invoices
Metered usage is invoiced every 30 days, emailed to you, and paid online from the Billing page.
Use cases
What teams run on it
01
First integration
Start with the quickstart and API reference.
02
Unexpected errors
Match the status code to the answers above.
03
Usage questions
Check the Usage and History pages.
04
Billing questions
Check the Billing page, then email our team.
How it works
From sign-up to production
- 1
Search these answers
Most questions are covered here.
- 2
Check the console
History, Usage and Billing show what happened.
- 3
Email our team
Include a request id, never your API key.
Control
Built so you stay in charge
Keys you control
API keys are shown once at creation and stored only as a hash. Revoke any key instantly from Settings.
Spend you can see
Every request carries its own receipt: tokens in, tokens out, latency and price. Nothing is billed that you cannot trace.
Limits you set
Per-account rate limits and a free monthly allowance keep experiments safe before they become production.
Data you decide on
You choose exactly what you send. Repeat prompts can be served from your cache, and cache hits are never billed.
For developers
One endpoint. Every model.
Send a prompt and a capability, or name a model. One bearer key, one small JSON body, and a full receipt on every response.
curl https://YOUR-DOMAIN/api/ainf/inference \
-H "Authorization: Bearer $AI_INFERENCE_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: order-1042" \
-d '{"prompt":"Summarise this ticket in one line","capability":"chat"}'Questions, answered
Why was a response not billed?
It was a cache hit, or it fell inside your free monthly allowance.
Why did I get a different model than last time?
If you sent a capability rather than a model_id, the router picks the lowest-priced eligible model and may fail over. Pin a model_id for consistency.
What does 502 mean?
No eligible model could complete the request after fail-over. Retry with backoff and the same Idempotency-Key.
How do I contact support?
Email our team with the request id and what you expected to happen.