FAQ

A credits-first AI answer API. You send a prompt; we classify the task, route to a model chain, and fail over across providers so your app keeps answering.

Not for the primary product path. Use POST /api/v1/answer with a prompt, or the OpenAI-compatible endpoint with model: "auto".

Each task type has an ordered provider/model chain. If the primary fails (error, timeout, circuit open, disabled), we try the next lane. The response includes attempts and whether fallback was used.

Buy prepaid credit packs ($10 / $25 / $50). 1 credit = $0.0001. Successful answers are charged at estimated provider token cost × 1.10, rounded up. There is no required monthly subscription.

Credit purchases are generally non-refundable except where required by applicable law. Unused credits remain on your account until used or the account is deleted. See Terms for details.

No. ElseLane does not retain prompt or answer bodies. We store usage metadata (status, model, provider, credits, latency) for billing, dashboards, and abuse prevention.

High-risk patterns (SSNs, payment cards, private keys, common cloud secrets) are blocked by default before any provider call. See Docs and Security.

Platform keys for providers such as Google, OpenAI, Anthropic, and others configured in routing chains. Your prompt is sent only to the providers tried for that request.

Create el_live_ or el_test_ keys in the dashboard. Send Authorization: Bearer …. Keys can be revoked anytime; last-used timestamps help you audit activity.

ElseLane aims for high availability but does not currently offer a contractual uptime SLA. Failover improves resilience; it cannot guarantee every request succeeds if all lanes fail.

ElseLane is a product of Boolean Array Canada (Ontario, Canada). Contact contact@elselane.com.