client-react: v1.11.0

1.11.0 (2026-10-29)

Features

• Added the includeExpressionEvents option to usePipecatConversation (RTVI 2.3.0+): each sound or pause the bot performs becomes a part of its message, with outputType "expression_sound" or "expression_pause", placed where the bot's speech had reached. It's off by default, and onMessageCreated and onMessageUpdated report messages without them. A reply that is only a sound isn't included, since it has no message of its own; RTVIEvent.BotExpression still reports it. Bots send sounds and pauses only to @pipecat-ai/client-js 1.16.0 or later. (9c70fdf)
• Added usePipecatBotDelivery(), which returns the bot's delivery now, such as ["cheerful"], or null while it isn't speaking (RTVI 2.3.0+). It requires @pipecat-ai/client-js 1.16.0 or later; with an older version it stays null. (9669cf5)

Bug Fixes

• Fixed spoken progress for a segment no message part holds, such as an expression sound's, moving the spoken text of the part being spoken in usePipecatConversation. (3c76005)
• usePipecatConversation now ends the assistant message when the bot is interrupted, so the bot's reply to text sent with sendText() and run_immediately appears as a new message instead of being appended to the one it cut off. Text the bot never got to say stays marked as unspoken. Requires a bot on RTVI protocol 2.0.0 or later. (68428ba)