OpenClaw Launches Active Memory Sub-Agent and Local MLX Voice Processing

OpenClawOpenClaw

· Updated

OpenClaw 2026.4.10 introduces an Active Memory plugin that uses a sub-agent to retrieve context before generating replies. This update also adds local voice processing for macOS via MLX and native support for the Codex model ecosystem.

OpenClaw, an open-source self-hosted AI assistant that runs across messaging platforms, released version 2026.4.10 featuring an optional Active Memory plugin. This plugin acts as a dedicated memory sub-agent that automatically pulls relevant preferences and past conversation details into the context window (the maximum data a model processes at once) before the assistant responds.

This release addresses the friction of manual context management by automating retrieval-augmented generation (grounding AI in external data) through a background agent loop. Adding local MLX-based speech for macOS reduces reliance on cloud-based voice APIs, moving toward a local-first architecture that prioritizes privacy and lower latency for multimodal interactions.

You can now manage tool execution through the new exec-policy CLI command and access native Codex features like managed authentication. The update is available as a free download on GitHub, with new integrations for Microsoft Teams actions and support for Seedance 2.0 video generation metadata.

View the full update on github.com
OpenClaw🦞
OpenClaw🦞
@openclaw
X

OpenClaw 2026.4.10 🦞 🧠 Active Memory plugin 🎙️ local MLX Talk mode 🤖 Codex app-server harness plugin 🧾 Teams pins/reactions/read actions 🛡️ SSRF hardening + launchd fixes stability, but with attitude🦞 https://t.co/PW7WDumTf1

55retweets535likes
View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from OpenClaw →

Share this update