Brain
Huawei Cloud's embodied world model, trained on Ascend compute, that uses a shared world-model backbone for prediction-enhanced world representation and a DualAct architecture to generate robot actions directly from the world model. Huawei Cloud announced it on November 16, 2026 and says model weights, inference and post-training code will open on November 30, with training-framework and Ascend compute code shared only with core partners.
World model · Maturity: Research · Closed
Machine-readable surfaces
- Markdown mirror: /brains/huawei-cloud-cloudrobo-wm1-2.md
- RSS feed: /brains/huawei-cloud-cloudrobo-wm1-2/feed.xml
- JSON-LD: embedded in this page’s head
- REST API: /v1/brains/dcc1a2ea-d561-4867-b977-f9a5c8fbaddc
- Revision history: /brains/huawei-cloud-cloudrobo-wm1-2/history
- Data documentation: /data
- Query this programmatically: Deploy MCP
Developed by (1)
Common questions
What is CloudRobo-WM1.2?
Who developed CloudRobo-WM1.2?
Is CloudRobo-WM1.2 open source?
What type of AI is CloudRobo-WM1.2?
What is CloudRobo-WM1.2's maturity stage?
Which robots run on CloudRobo-WM1.2?
Sources (2)
Methodology: Verified · 2 sources (no primary) · last reviewed 2026-11-18
Verification posture
Verified
Low confidence
Review state
Stable
Last reviewed 2026-11-18
Maturity + lifecycle
Maturity stage: research
Sources by quality tier
- 2
- unclassified
- Unclassified source
The framework is documented at /methodology. Corrections at /corrections. Reviewer: DEPLOY editorial team.
Methodology surface for CloudRobo-WM1.2.Canonical ID dcc1a2ea-d561-4867-b977-f9a5c8fbaddc