OpenClaw v2026.4.25 (Beta) — Full TTS Overhaul: ElevenLabs v3, Azure Speech, Auto-TTS Controls, and OTEL Expansion
OpenClaw’s voice capabilities just got a serious upgrade. Released April 26, v2026.4.25-beta.4 ships what the changelog calls a “complete TTS overhaul” — and the scope of the changes justifies that framing. This isn’t a new provider added here or a bug fix there. It’s a ground-up rework of how OpenClaw thinks about, routes, and controls text-to-speech across agents, personas, and sessions. Install it now with: npm i [email protected] What’s New: TTS Architecture Auto-TTS Controls The headline feature is auto-TTS: agents can now speak their responses automatically without requiring the user to explicitly invoke /tts. Configuration is flexible — you can toggle it per agent, per account, or at the session level. Duplicate suppression is built in, so agents that respond to the same message in multiple ways won’t read out every variant aloud. ...