Dyna-2.1 by Dyna Robotics: status and timeline

Milestones

Next · Hundreds of robots deployed by mid-2027

Dyna Robotics foundedSep 2024Complete.
Seed funding round of $23.5M led by CRV and First Round25 Mar 2025Complete.
Dyna-2.1 deployments begin in fulfillment, hotels and restaurantsJun 2025Complete.
Series A funding of $120M led by RoboStrategy, CRV and First Round15 Sep 2025Complete.
Dyna-2.1 launched commercially on September 29, 202629 Sep 2026Complete.
Hundreds of robots deployed by mid-2027Target Jun 2027Not yet reached.

Upcoming

Q4 2026Commercial deployments expand to 100+ units (next)
Jun 2027Hundreds of robots deployed by mid-2027
Jun 2027Hundreds of robots in operation; Series B funding

Current obstacles

Generalizing across domainsExtending reliable operation beyond fulfillment, hospitality and food service to other commercial tasks requires additional training and validation.
Scaling manufacturing and costsRamp production to hundreds by mid-2027 while reducing unit costs and maintaining reliability.

Physics limits

Wheeled bases have traction limitsWheeled robots cannot traverse stairs or extreme slopes without additional mechanisms, limiting deployment locations compared to legged robots.
Dexterous manipulation is computationally demandingRunning vision-language models and reinforcement-learned policies for complex manipulation requires significant onboard or edge compute, increasing cost and power consumption.

How it works

4 parts

Wheeled base

Four-wheel platform for stable mobility

Four steerable wheels enable stable operation on varied surfaces without balance adjustment.

Dual arms

Configurable grippers and hands

Swappable end-effectors: parallel-jaw grippers for pick-and-place or dexterous hands for complex manipulation.

Multi-sensory feedback

Vision, tactile and proprioceptive sensing

Vision-language models, tactile sensors and joint sensing enable perception of environment and self-state.

Foundation models

Proprietary policy and world-action modeling

Dyna-2 policy layer and world-action models reason about task sequences and adapt to novel conditions.

Spec sheet

Spec	Dyna-2.1
Fulfillment throughput	200 units/hour^R (reported)
Placement accuracy	±2 mm^R (reported)
Hotel room turnover time	22 minutes^R (reported)
First-pass hotel inspection rate	97%+^R (reported)
Food service throughput	100 dishes/hour^R (reported)
Time per dish	<40 seconds^R (reported)

^R reported by the company

Papers & demos

1 item

2024demo
Dyna-2.1 prototype unveiled (opens dyna.co)Introduction of semi-humanoid robot for commercial task automation.

About Dyna Robotics

The team behind Dyna-2.1

Dyna Robotics

Dyna-2.1 robot, foundation models, manipulation

Dyna Robotics develops Dyna-2.1, a semi-humanoid robot for commercial dexterity tasks. Founded 2024, raised $140M+, deployed in fulfillment, hospitality and food service.

Dyna-2.1: 200 units/hour at ±2mm accuracy; 22-min hotel room turnover; 100 dishes/hour service.
Raised $23.5M seed (March 2025) and $120M Series A (September 2025) from CRV, First Round, RoboStrategy, NVentures, Amazon, Samsung and LG.
Launched commercially September 2026 with deployments in hotels, restaurants and laundromats.

dyna.co (opens dyna.co)Company profile →

Founded: 20242 yrs
Headquarters: United States
Status: Private
Valuation: $600MprivateSep 2025
Raised: $144M2 rounds
Last round: Series A · $120M15 Sep 2025
Works in: RoboticsRobot brains
Coverage: 1 program · 3 updateslatest 12d agochecked 10 Oct
Lead investors: RoboStrategyCRVFirst Round Capital
Investors: RoboStrategy (opens robostrategy.com)NVIDIA (opens nvidia.com)Amazon (opens amazon.com)Samsung (opens samsung.com)LG (opens lg.com)
People: Lindon Gao (opens dyna.co)Co-founder & CEOYork Yang (opens dyna.co)Co-founderJason Ma (opens dyna.co)Co-founder, former DeepMind scientist

Spec

Dyna-2.1

Fulfillment throughput

200 units/hour^R (reported)

Placement accuracy

±2 mm^R (reported)

Hotel room turnover time

22 minutes^R (reported)

First-pass hotel inspection rate

97%+^R (reported)

Food service throughput

100 dishes/hour^R (reported)

Time per dish

<40 seconds^R (reported)