Dexterity

aid: dexterity
name: Dexterity
description: >-
  Dexterity is a Redwood City, California Physical AI company founded in 2017
  by Samir Menon and a group of Stanford roboticists. It builds a full-stack
  industrial robotics platform that combines the Mech — a roving, dual-arm
  superhumanoid robot — with Foresight, an in-house world model trained on
  more than 100 million autonomous actions in production. Foresight
  orchestrates 68+ skill agents (packing, motion, actualization, trajectory,
  force-control) across the IRIS hardware abstraction layer to drive Kawasaki
  arms, HiWin motion components, and Dexterity's own mobile bases at sub-400ms
  decision latency with zero reported safety incidents. Customers include
  FedEx, UPS, Sagawa Express, Maersk, and VFC, and the company has raised more
  than 291M USD from Kleiner Perkins, Lightspeed Venture Partners, Obvious
  Ventures, Sumitomo Corporation, and B37 Ventures. Dexterity's commercial
  offering is enterprise Physical AI deployment (no self-service developer
  API), but it does run a public, research-grade REST API — the Foresight
  Packing Challenge — that exposes the bin-packing facet of Foresight as a
  programmable sequential placement game, with a 50,000 USD grand prize and
  participation restricted to .edu email holders.
type: Index
position: Provider
access: 3rd-Party
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
tags:
  - Physical AI
  - Industrial Robotics
  - Robotics
  - Warehouse Automation
  - Logistics
  - Manufacturing
  - World Model
  - Foresight
  - Mech
  - Dual-Arm
  - Truck Loading
  - Palletizing
  - Depalletizing
  - Singulation
  - Research API
  - Packing Challenge
url: https://raw.githubusercontent.com/api-evangelist/dexterity/refs/heads/main/apis.yml
created: '2026-05-24'
modified: '2026-05-24'
specificationVersion: '0.20'
apis:
  - aid: dexterity:dexterity-foresight-packing-challenge
    name: Dexterity Foresight Packing Challenge API
    description: >-
      Public REST API exposing the Foresight Packing Challenge as a sequential
      3D bin-packing problem. Autonomous agents start a game, place boxes
      sequentially inside a 2.0m x 2.6m x 2.75m truck, and submit final
      packing density scores to a public leaderboard. Two modes are supported:
      dev (unlimited, physics-free testing) and compete (full physics
      simulation with a 50 games/day per-API-key cap). API keys are issued by
      reaching at least 50% density in the public Truck Loading Game; full
      leaderboard participation and the 50,000 USD grand prize are restricted
      to .edu (or equivalent university) email holders. Interactive Swagger
      docs live at /challenge/api/docs.
    humanURL: https://dexterity.ai/challenge
    baseURL: https://dexterity.ai/challenge/api
    tags:
      - Physical AI
      - Foresight
      - Packing Challenge
      - REST API
      - Research API
      - Bin Packing
      - Leaderboard
    properties:
      - type: Documentation
        url: https://dexterity.ai/challenge
      - type: APIReference
        url: https://dexterity.ai/challenge/api/docs
      - type: GettingStarted
        url: https://dexterity.ai/play
      - type: Authentication
        url: https://dexterity.ai/challenge
      - type: OpenAPI
        url: openapi/dexterity-foresight-packing-challenge-openapi.yml
      - type: JSONSchema
        url: json-schema/dexterity-foresight-game-schema.json
      - type: JSONSchema
        url: json-schema/dexterity-foresight-box-schema.json
      - type: JSONSchema
        url: json-schema/dexterity-foresight-placed-box-schema.json
      - type: JSONSchema
        url: json-schema/dexterity-foresight-truck-config-schema.json
      - type: JSONSchema
        url: json-schema/dexterity-foresight-leaderboard-entry-schema.json
      - type: JSONStructure
        url: json-structure/dexterity-foresight-game-structure.json
      - type: Example
        url: examples/dexterity-foresight-start-game-example.json
      - type: Example
        url: examples/dexterity-foresight-place-box-example.json
      - type: Example
        url: examples/dexterity-foresight-stop-game-example.json
      - type: Example
        url: examples/dexterity-foresight-get-status-example.json
      - type: Example
        url: examples/dexterity-foresight-my-games-example.json
      - type: Example
        url: examples/dexterity-foresight-display-name-example.json
      - type: Example
        url: examples/dexterity-foresight-leaderboard-example.json
      - type: Example
        url: examples/dexterity-foresight-health-example.json
      - type: NaftikoCapability
        url: capabilities/packing-game-play.yaml
      - type: NaftikoCapability
        url: capabilities/leaderboard-tracking.yaml
common:
  - type: Website
    url: https://www.dexterity.ai
  - type: Platform
    url: https://www.dexterity.ai/platform
  - type: Foresight
    url: https://www.dexterity.ai/blog/foresight
  - type: Mech
    url: https://www.dexterity.ai/platform
  - type: Challenge
    url: https://dexterity.ai/challenge
  - type: TruckLoadingGame
    url: https://dexterity.ai/play
  - type: About
    url: https://www.dexterity.ai/about
  - type: Blog
    url: https://www.dexterity.ai/blog
  - type: Contact
    url: https://www.dexterity.ai/contact
  - type: TermsOfService
    url: https://dexterity.ai/terms-conditions
  - type: LinkedIn
    url: https://www.linkedin.com/company/dexterityinc
  - type: SpectralRules
    url: rules/dexterity-rules.yml
  - type: Vocabulary
    url: vocabulary/dexterity-vocabulary.yml
  - type: JSON-LD
    url: json-ld/dexterity-context.jsonld
  - type: Plans
    url: plans/dexterity-plans-pricing.yml
  - type: RateLimits
    url: rate-limits/dexterity-rate-limits.yml
  - type: FinOps
    url: finops/dexterity-finops.yml
  - type: Features
    data:
      - name: Foresight World Model
        description: In-house world model trained on 100M+ autonomous actions for spatial reasoning, prediction, and dual-arm orchestration.
      - name: 68+ Autonomous Skill Agents
        description: Specialized agents including Packing, Motion, Actualization, Trajectory, and Force-Control coordinated in real time.
      - name: IRIS Hardware Abstraction Layer
        description: Unified API enabling deployment of Physical AI across Kawasaki arms, HiWin motion components, and Dexterity-built mobile bases.
      - name: Mech Dual-Arm Robot
        description: Roving superhumanoid robot with two arms designed for industrial logistics tasks.
      - name: Sub-400ms Decision Loop
        description: End-to-end perception-to-action loop under 400ms in production environments.
      - name: Zero Safety Incidents
        description: 100M+ autonomous decisions in production with no reported safety incidents to date.
      - name: Foresight Packing Challenge API
        description: Public, research-grade REST API exposing the bin-packing facet of Foresight with a public leaderboard and 50,000 USD prize.
  - type: UseCases
    data:
      - name: Trailer Loading
        description: Autonomous dual-arm loading of mixed-SKU boxes into truck trailers at the dock.
      - name: Trailer Unloading
        description: Autonomous depalletization and unloading of inbound trailers.
      - name: Palletizing
        description: Building stable pallets from heterogeneous case mixes.
      - name: Depalletizing
        description: Breaking down pallets into individual cases for downstream handling.
      - name: Singulation
        description: Separating items from bulk flow into a single, oriented stream for sortation.
      - name: Aircraft Loading and Unloading
        description: Cargo loading and unloading for air freight operations.
      - name: Bin Packing Research
        description: Academic and industry research into sequential 3D bin packing using the public Foresight Packing Challenge API.
  - type: Integrations
    data:
      - name: Kawasaki Heavy Industries
        description: Partnership producing intelligent industrial robot arms for warehouse applications.
      - name: HiWin
        description: Precision motion control components integrated into Dexterity systems.
      - name: NVIDIA
        description: GPU acceleration of Foresight via TensorRT and CUDA delivering a 17x speedup.
      - name: FedEx
        description: Production deployment partner; Dexterity highlighted at FedEx Investor Day 2026.
      - name: Sagawa Express
        description: Production logistics deployment in Japan via partnership with Sumitomo Corporation.
      - name: Sumitomo Corporation
        description: Japan market partner and investor; led recent funding rounds.
      - name: UPS
        description: Enterprise customer deploying Dexterity robots in production logistics operations.
      - name: Maersk
        description: Global logistics customer running Dexterity systems.
      - name: VFC
        description: Enterprise customer.
  - type: Solutions
    data:
      - name: Logistics Automation
        description: End-to-end automation of trailer-, pallet-, and case-handling tasks in warehouse and parcel operations.
      - name: Manufacturing Material Handling
        description: Dual-arm material handling on the manufacturing floor.
      - name: Air Cargo Handling
        description: Loading and unloading cargo for air freight workflows.
      - name: Research Access (Foresight Packing Challenge)
        description: Free research access for academic teams to a slice of Foresight via a public REST API and competition.
maintainers:
  - FN: Kin Lane
    email: [email protected]