Definition
Full definition of latency
A workflow's end-to-end latency includes: trigger detection (webhook ~1s, polling 1-15 min), queue time (~100ms), step execution time (varies by action: AI calls 2-10s, simple writes ~500ms), downstream propagation (~1s). Most well-designed workflows complete in 2-10 seconds. Tiny Command shows total latency in every run-log.
In practice
Latency examples
Typical workflow latency
Webhook (1s) + Queue (0.1s) + Action (1-5s) + Network (1s) = 3-7s total
Used by
Apps that exemplify latency
See latency in action across real integrations.
FAQ
Common questions about latency
How do I reduce workflow latency?
Use webhook triggers (not polling). Parallel-fan-out independent steps. Use faster AI models (Haiku > Sonnet > Opus). Skip unnecessary lookups.
When does latency matter most?
User-facing workflows (form submission → confirmation email) where humans wait. Async batch flows (nightly reports) tolerate higher latency.
