OpenClaw 2026.4.24 🦞 ☎️ Voice calls can now reach the full agent 🧠 DeepSeek V4 Flash + Pro join the team 🖱️ Browser automation got coordinate clicks + better recovery 🔧 Telegram, Slack, MCP, sessions, and TTS fixes More reach. Less duct tape. https://t.co/Oa9G9WpO84
OpenClaw Connects Real Time Voice to Full Agent Tools and Google Meet
OpenClaw, an open-source self-hosted AI assistant that connects models to messaging and automation, released version 2026.4.24 with a real-time agent consultation feature. This allows voice calls and the new Google Meet plugin to use fast voice loops while still accessing the full agent for tool-backed answers. It also integrates DeepSeek V4 Flash and Pro.
This update solves the latency trade-off in voice-based agents by separating the real-time audio bridge from the heavy reasoning layer. By adding a Google Meet participant plugin, OpenClaw moves beyond simple chat interfaces into active meeting orchestration, following a rapid release cycle that recently added GPT-5.5 support and multimodal xAI tools.
You can now deploy OpenClaw as a meeting participant to generate smart notes and track attendance. The update is available on GitHub, with deepseek-v4-flash as the default onboarding model. Developers can also leverage new OpenTelemetry spans to monitor tool execution and system performance.
OpenClaw🦞
@openclaw
55retweets533likes
View on XStill wondering? A few quick answers below.
The Google Meet plugin is a bundled participant tool that allows the AI assistant to join meetings. It supports personal Google authentication, real-time sessions via Chrome or Twilio, and can export meeting artifacts like attendance records and smart notes. It also includes recovery tools to inspect already-open meeting tabs without creating duplicates.
This feature allows real-time voice loops in phone calls or Google Meet sessions to hand off complex questions to the full OpenClaw agent. While the voice model handles fast, low-latency dialogue, it can consult the main agent logic to access specialized tools and provide deeper, data-backed answers without interrupting the live audio conversation.
OpenClaw has added native support for DeepSeek V4 Flash and DeepSeek V4 Pro to its bundled model catalog. DeepSeek V4 Flash is now set as the default model for new user onboarding. The update also includes specific fixes for how these models handle thinking and replay behavior during multi-turn tool interactions.
