OpenClaw 2026.4.5 Brings Built-In Video and Music Generation

OpenClaw 2026.4.5 landed April 6, 2026, marking one of the release’s most significant feature jumps. Built-in video generation now lets agents create videos directly through bundled providers (xAI grok-imagine-video, Alibaba Wan, Runway) and return the media in-reply. Music generation is equally native — bundled Google Lyria and MiniMax providers, plus workflow-backed Comfy support with async task tracking so finished audio delivers when ready.

The /dreaming feature — OpenClaw’s experimental memory consolidation system — is now generally available, with weighted short-term recall promotion across three cooperative phases (light, deep, REM), independent schedules, and recovery behavior. Dreams write to dreams.md instead of daily memory notes, and a Dream Diary surface is added to the Dreams UI.

Structured task progress brings step-by-step progress events for compatible UIs, and prompt-cache reuse improvements should cut latency by reducing redundant context on follow-up turns. The Control UI adds localization for Simplified Chinese, Traditional Chinese, Brazilian Portuguese, German, Spanish, Japanese, Korean, French, Turkish, Indonesian, Polish, and Ukrainian — bringing total language support to over 20.

Also notable: OpenClaw notes that Anthropic access was cut off while GPT-5.4 performance improved, prompting a shift in default provider usage for the platform.

Story of the Day. Video and music generation native to the agentic harness is a meaningful step toward closing the gap between “AI that talks” and “AI that creates” — particularly in an open-source stack.