Pay-as-you-go

Pay-as-you-go lets usage exceed your plan's monthly quota instead of hard-failing at the limit. When enabled, overage is metered and billed per operation.

Rates

Operation	Overage price
Memory retained	$0.01 each
Recall	$0.01 each
Reflect	$0.05 each

Usage within your plan's quota is always included — overage rates apply only beyond it. See Plans & Pricing for included quotas.

How metering works

Every billable operation records a usage event with its operation type, the bank/space, and a token estimate — the same numbers you see in Settings → Usage. Recall responses also return a usage object (tokens_used, max_tokens, truncated) so you can track cost client-side.

Enabling / disabling

Toggle pay-as-you-go in Settings → Billing:

• Off (default): requests that exceed your quota return 429 Too Many Requests. Upgrade your plan or wait for the monthly reset.
• On: requests continue past the quota and overage is added to your next invoice.

Rate limits vs quotas

• Quota — total operations per month (per plan). Overage is where pay-as-you-go applies.
• Rate limit — requests per minute (per plan). This is a burst guard and is not affected by pay-as-you-go; a 429 from the rate limit means slow down, not "out of quota."