How Fast Is the sipi.bot Decision?
About 5 milliseconds per evaluation — deterministic rules, no model in the path, no network round-trip to an LLM.
Why it's fast
Rules are deterministic comparisons, not inference.
The evaluate endpoint is a single lightweight call.
What the number means
Spend actions add ~5 ms — invisible next to API latency.
It's fast enough for payment rails and voice agents.
FAQ
Is it faster than a model-based check?
Yes — orders of magnitude. Rules don't reason, they decide.
Where does latency come from?
Your network to the endpoint — the evaluation itself is microseconds.
Related
Stop the next $12,400 night.
One API call (or MCP tool) in front of every agent transaction — APPROVED, BLOCKED, or FLAGGED, deterministic, ~5 ms, fully logged.
See plans — from $99/mo Try a live check