Steps
Speak
Announce a message on media players
Then step — What happens. Step type then-speak.
How it works
Speaks a text on one or more media players, optionally at a set volume, then restores what was playing. Announcements take turns on a player: one still speaking finishes first (tick Interrupt to cut in). Placeholders fill in live values: {{trigger.name}}, {{trigger.friendlyState}}, {{trigger.oldState}}, {{trigger.attribute.battery}}, {{var.x}}.
Example
Device changes (Front door) → Speak (Kitchen speaker, "{{trigger.name}} is {{trigger.friendlyState}}")A spoken announcement in the kitchen.
Options
| Option | Key | Kind | Required | Notes |
|---|---|---|---|---|
| Players | entityIds | entities | yes | domains: media_player |
| Say | text | text | yes | supports {{…}} placeholders; multi-line |
| Volume | volume | number | no | range 0–100 % |
| Resume what was playing afterwards (Sonos) | restoreAfter | boolean | no | — |
| Interrupt an announcement already playing (otherwise wait for it to finish) | interrupt | boolean | no | — |
Wire format
What the step stores in the flow definition's data bag (JSON schema).