FAQ

Does the gateway store my prompts?

By default, no. IQ Routing stores request metadata (timestamps, model and provider, token counts, latency, cost, and the routing decision), which the owning team can browse in the requests browser. Prompt and completion payloads are not logged by default. Turning on fuller payload capture for debugging is not a self-serve dashboard toggle today -- contact us to enable it for your organization -- and when it is on, payloads are visible only to authorized team members and are kept for your org's configured retention window.

Does IQ Routing offer Zero Data Retention (ZDR)?

Yes. ZDR is an org-level setting you turn on in Settings, and with it on, request and response bodies are never persisted and payload capture cannot be armed while it is on. Operational metadata is still retained for billing and the requests browser. When loop detection is enabled, we also retain secret-key fingerprints, activity counts, loop flags and estimated wasted cost to detect agents stuck in repeated loops. See /docs/security for retention details.

Do you train on my prompts?

Not through any automated pipeline, and not by default for any customer. IQ Routing keeps an internal, manually-run export script an engineer can use to build training data for the systems that select a model for each request, but only for an organization that has explicitly opted in, or for a design partner -- never for any other organization. For an org in that opted-in or design-partner set, the script only draws from payloads that org has retained, and there is currently no per-request control to exclude a specific retained row from the export; not opting in is the org-level control. Retention itself is a separate, org-level setting: every org runs metadata-only by default, so no prompt or completion text is written, and turning on fuller payload capture for debugging is not a self-serve dashboard toggle today -- contact us to enable it for your organization. The script is not scheduled and does not run automatically. Not opting in, staying on the metadata-only default, or turning on Zero Data Retention all keep your prompts out of it entirely.

How do I tell which rate limit a 429 hit?

Every 429 response carries an X-RateLimit-Scope header naming the limit that fired. The scope values are rps, rpm, tpm, daily_budget, monthly_budget, and key_spend_ceiling, so a client can back off on a burst and escalate on a blown monthly budget without opening a support ticket. Retry-After and X-RateLimit-Reset ride along on the rps, rpm, and tpm windows; rps uses a one-second window while rpm and tpm use the sixty-second window, so read Retry-After rather than assuming a minute. A spend cap has no rolling reset, so do not assume those two headers are always present.

Which models are supported?

IQ Routing supports a broad set of providers. See /docs/providers for the full provider list and /docs/models for the model list. Routing aliases keep up with provider roster changes, so a roster refresh does not change your call site. Naming a specific model keeps the request with that model's provider, including claude-fable-5-1, Anthropic's current Fable model; see /docs/routing for details, including the older claude-fable-5 id.

Can I bring my own provider keys?

Yes, and you should. Per-org provider credentials live in /providers, encrypted at rest. The gateway uses your keys for upstream calls. Your providers bill model usage separately. See /docs/providers for the full provider roster and setup details.

Is there a free tier?

Free is $0 and needs no card. New Free accounts get 60 requests/minute and 1,000 requests/day for their first 7 days, then standard Free limits. Token and spending limits also apply. Your connected providers bill model usage separately; Free includes no provider credit.

How do I correlate a gateway error with a request?

Every response includes an X-Request-Id header. Open /requests/<id> to see the full pipeline trace.