OpenClaw v2026.5.4: Google Meet Voice Agents Get Gemini Realtime Bridge

OpenClaw released v2026.5.4 today, the first May release, bringing significant improvements to Google Meet and voice call integrations.

Gemini Realtime Voice Bridge for Meet

The headline feature is a re-architected Twilio dial-in flow for Google Meet. When Meet participants dial in via Twilio, their audio now flows through OpenClaw’s realtime Gemini voice bridge with paced audio streaming, backpressure-aware buffering, and barge-in queue clearing. Notably, there’s no TwiML fallback during realtime speech, which gives Meet participants a much snappier OpenClaw voice agent.

Other Changes in 2026.5.4

  • Windows localhost fix: Gateway now binds only to 127.0.0.1 on Windows, preventing libuv’s dual-stack ::1 behavior from wedging localhost HTTP requests.
  • Plugin migration: Configs referencing uninstalled official plugins now get install hints pointing to openclaw plugins install rather than telling operators to remove valid plugin config.
  • OpenAI/Codex media: Codex audio transcription is now advertised in runtime and manifest metadata; active Codex chat models route to the OpenAI transcription default instead of sending chat model IDs to audio transcription.
  • Performance: Agents and plugins received workspace-scoped metadata improvements to reduce cold plugin metadata scans.

Full changelog: github.com/openclaw/openclaw/releases

← Back to News