CB Insights - The physical AI models market map: Behind the arms race to control robot intelligence - January 2026
FreeBehind the arms race to control robot intelligence
FreeFree tier
About CB Insights - The physical AI models market map: Behind the arms race to control robot intelligence - January 2026
This research report from CB Insights maps the physical AI models landscape, covering over 70 companies across 10 categories including data simulation, model approaches (VLMs, VLAs, world models), foundation models, and observability. It highlights key takeaways such as the importance of proprietary training data, the surge in world model investment ($6.9B in 2025), and the gap in multi-robot coordination. The report is based on CB Insights' Mosaic scores and data, with an accompanying book of scouting reports.
Key Features
Identifies 70+ companies across 10 physical AI model categories
Includes Mosaic health scores for company sorting
Covers data simulation, VLMs, VLAs, world models, foundation models, observability
Highlights proprietary training data as critical competitive advantage
Notes record $40.7B in robotics venture funding in 2025 (up 74% YoY)
Provides book of scouting reports for CB Insights customers
Pros & Cons
Pros
- Comprehensive market map with over 70 companies across multiple categories
- Data-driven insights using Mosaic scores for company health
- Covers critical model architectures (VLMs, VLAs, world models) and data simulation
- Highlights competitive dynamics around proprietary training data and multi-robot coordination
Cons
- Map is not exhaustive; may exclude smaller companies or those with lower Mosaic scores
- Full access to company details and scouting reports requires CB Insights subscription
- Focus primarily on US and Chinese companies; global representation may be limited
Best For
Market research for investors tracking physical AI trendsCompetitive analysis for robotics and AI companiesStrategic planning for startups in the physical AI spaceUnderstanding key players, investment dynamics, and technology categories
FAQ
What is physical AI?
Physical AI enables robots to operate in the physical world and learn from data rather than pre-programmed rules.
How many companies are in the market map?
The map identifies over 70 companies across 10 categories.
What is Mosaic score?
Mosaic predicts the health and success likelihood of private companies, used to sort companies in the map.
Is the map exhaustive?
No, the map is not exhaustive. It includes leading public corporations and private companies with Mosaic scores above 500; markets with 5 or fewer companies include all companies.