FAQ

For everyone

Is Keln OpenAI-compatible? Yes, /v1/chat/completions and /v1/models, streaming and non-streaming, with standard SDKs. Change the base URL and keep your code. Quickstart →

Which provider served my request? Not exposed. Keln routes across a pool of more than 100 suppliers under one price and one SLA per model; the pool is continuously verified. Why →

Do you store my prompts? No. Not in logs, not for analytics, not for verification. Zero data retention →

For users

How do I pay? Workspaces are prepaid: top up by card through Stripe checkout, $5 minimum, with the card processing fee (2.9% + $0.30) itemized at checkout. Your card is saved automatically, so later top-ups are one click, and auto top-up keeps the balance from running out under sustained traffic. Billing →

Can I get a refund? Refunds are handled by support, case by case, and can only cover credits you haven't spent; refunded credits come off your balance. Billing →

How do I control spend? Prepaid balance per workspace + optional per-key budgets with daily/weekly/monthly resets. A request over budget fails with a 402; usage is visible on the Usage page, and low-balance email alerts are automatic. Keys & budgets →

Can my team share billing? Yes, teams have their own wallet, keys, and roles, with attribution at the key level: every key shows who created it, and usage breaks down per key. Teams →

Can I try models without writing code? The portal Playground runs every catalog model against the live API, with latency and token metrics per response. It uses your real API key and balance, so a brand-new account needs a key and some credit first, the playground walks you through both.