# Generalist

> Generalist is a San Mateo robotics AI company developing foundation models that let robot arms learn dexterous physical tasks from experience and short demonstrations. Its GEN-1 family targets adaptable automation across robot bodies and tools; reported results are strongest on short, controlled tasks, with commercial access offered through partnerships.

- **Founded:** 2024
- **Headquarters:** San Mateo, United States
- **Founders:** Pete Florence (Co-founder and CEO), Andy Zeng (Co-founder and chief scientist), Andrew Barry (Co-founder and CTO)
- **Products:** GEN-0, GEN-1, GEN-1.5, Wearable data collection devices
- **Notable:** GEN-0 reported pretraining on more than 270,000 hours of real manipulation data in November 2025., GEN-1 reported a dataset of more than 500,000 hours of physical interaction data and at least 99% success on several selected simple tasks in April 2026., GEN-1.5 reported 59% average one-shot success across 10 short tasks and 83% after few-shot adaptation on five minutes of data per task in August 2026.

## Products & services

- **GEN-0** — Embodied foundation model trained on large-scale real physical interaction data; introduced scaling and transfer results.
- **GEN-1** — Robot learning system that Generalist reports reached at least 99% success on several selected simple tasks with roughly one hour of robot data per task. Early access is offered to partners.
- **GEN-1.5** — Embodied foundation model that uses brief sensorimotor demonstrations as prompts for short tasks and can adapt further with a few minutes of task data.
- **Wearable data collection devices** — In-house devices capture human physical interaction data for model pretraining; the company has not published them as a customer product.

## Achievements

- GEN-0 reported pretraining on more than 270,000 hours of real manipulation data in November 2025.
- GEN-1 reported a dataset of more than 500,000 hours of physical interaction data and at least 99% success on several selected simple tasks in April 2026.
- GEN-1.5 reported 59% average one-shot success across 10 short tasks and 83% after few-shot adaptation on five minutes of data per task in August 2026.
- Generalist reported roughly 9,000 end-effector variations in its GEN-1 training data in July 2026.

## Latest updates

- **2026-08** — Generalist published GEN-1.5 one-shot and few-shot task results.
- **2026-08** — TechCrunch reported a near-$200 million Series B extension and a $3 billion valuation; Generalist did not confirm the report.
- **2026-07** — Company demonstrated GEN-1 across a wide variety of robot end effectors.
- **2026-06** — Company announced $400 million in new funding, taking announced total funding above $500 million.

## Links

- Website: https://generalistai.com/
- LinkedIn: https://www.linkedin.com/company/generalistai/
- Twitter/X: https://x.com/GeneralistAI
- YouTube: https://www.youtube.com/@Generalist_AI

---

Profile page: https://yespress.io/generalist
Published by YesPress — https://yespress.io
Last updated: 2026-09-29