RLWRLD
Robot foundation-model company training five-finger robot hands for factory work with its RLDX models and in-house motion-capture data.
Headquartered in San Francisco, United States.
RLWRLD builds robot foundation models aimed at dexterous hand work in industry. Its flagship model family is RLDX; the company introduced RLDX-1 on its blog on April 29, 2026 as a dexterity-first foundation model for robot hands, and says the model can be fine-tuned for bodies ranging from dexterous hands to full humanoids. RLWRLD says it trains on proprietary 4D+ motion-capture data collected on factory floors rather than in simulation, and that it works with enterprise customers in Japan and Korea under three-to-five-year commitments. Its website lists a US headquarters in San Francisco and offices in Seoul and Tokyo. RLWRLD researchers are co-authors, with POSTECH, KAIST and Hanyang University, of the PointWAM preprint on 3D world action models for dexterous manipulation posted in October 2026.
- Latest funding
- —
- Total funding
- —
- Latest valuation
- —
- Founded
- —
- Country
- United States
- PointWAM: 3D World Action Modeling for Dexterous Robotic ManipulationPOSTECH, KAIST, Hanyang University and RLWRLD · Oct 2, 2026