2026-06-17 👁️ — views

Humanoid Robot Race Index — Tesla Optimus, Figure AI, Agility, 1X, Boston Dynamics: Mid-2026 Benchmark

Tesla Optimus, Figure 02, Agility Digit, 1X NEO, and Boston Dynamics Atlas benchmarked on funding, production status, and 2026 deployment targets.

Why this benchmark exists

Humanoid robots crossed a threshold in 2023–2025: they moved from research platforms to factory-floor pilots. Investment surged, incumbents like Amazon and BMW signed deployment agreements, and two new well-funded startups — Figure AI and 1X Technologies — emerged to challenge Boston Dynamics’ decade-long lead. This index benchmarks the five most significant programs on the metrics that matter for the industrial ramp: capital raised, production status, deployment scope, and the credible timeline to volume.

Humanoid Robot Race Index

Tesla Optimus

Tesla’s humanoid program is the most vertically integrated of the five. Optimus shares the same vision-only neural network stack as FSD — cameras feed directly into a learned policy without LIDAR or HD maps. This is intentional: Tesla believes a general-purpose robot must navigate unstructured environments the way humans do.

Tesla’s key structural advantage is data. Its fleet of more than 6 million FSD-equipped vehicles generates enormous quantities of real-world video for training vision models, and that same foundation transfers (partially) to robot perception. The challenge is manipulation: driving data does not teach hands to grasp, sort, or assemble parts. Tesla has built a dedicated robot-simulation pipeline and a dexterous hand with 22 degrees of freedom per hand to address this.

Musk has stated publicly that Optimus could eventually exceed Tesla’s automotive revenue. That claim depends on hitting volume targets that have repeatedly slipped, and on solving dexterous manipulation at scale — an open research problem as of mid-2026.

Figure AI

Figure AI was founded in 2022 by Brett Adcock (formerly of Archer Aviation) and has raised a total of more than $900 million through mid-2026, including a landmark $675 million Series B in early 2024 backed by OpenAI, Microsoft, NVIDIA, Jeff Bezos, and Intel Capital. That round validated Figure as the most well-capitalized independent humanoid startup in the world.

Figure’s AI architecture is the most language-model-centric of the group: Figure 02 uses an OpenAI-developed multimodal model that takes camera images and spoken instructions as input and outputs robot actions. In a widely shared demo, an operator gave verbal instructions and Figure 02 correctly sorted objects, cleaned a table, and identified items — all in natural language. The practical implication is that Figure robots can be reprogrammed for new tasks by updating the language model policy, not by rewriting low-level motion code.

The BMW Spartanburg pilot is the program’s key commercial proof point. BMW’s South Carolina plant is the largest BMW factory in the world by output; if Figure can reliably perform body-shop tasks there, it unlocks a strong OEM reference case.

Agility Robotics (Amazon)

Agility Robotics is the oldest commercial humanoid company on this list. Founded in 2015 as a spinout from Oregon State University, it introduced Digit — a bipedal robot with a torso, arms, and a head sensor cluster — in 2019. Amazon invested in Agility in 2023 and became its most significant commercial partner, deploying Digit at its Sequoia facility.

Digit is the most commercially advanced robot in this index by one measure: it is the only one operating inside a live major-retailer fulfillment network under normal commercial conditions (not a research lab or controlled pilot). The Sequoia facility was designed with Digit in mind — charging docks, standardized container heights, and workflow software built around robot-human collaboration.

Agility opened RoboFab in Salem, Oregon in 2023 — the first purpose-built humanoid robot factory in the US — and targets a capacity of up to 10,000 units per year. That capacity would make Digit the highest-volume humanoid in commercial deployment if fully utilized.

1X Technologies

1X Technologies (formerly Halodi Robotics) is a Norwegian startup that has attracted backing from OpenAI and other investors. It operates two robot platforms: EVE (a wheeled, commercial-duty robot) and NEO (a fully bipedal humanoid intended for home and general-purpose use).

1X’s differentiation is its actuator philosophy: EVE and NEO use soft, compliant actuators that limit peak force output. This makes the robots less likely to injure a human in a collision — a meaningful property for home deployment where unpredictable obstacles (children, pets, clutter) are the norm. The trade-off is lower payload capacity relative to Boston Dynamics or Tesla Optimus, which use rigid high-torque actuators.

NEO is the program to watch: it targets the home assistant market, which is orders of magnitude larger than the factory market but also far harder to navigate from a reliability, safety, and regulatory perspective.

Boston Dynamics (Hyundai)

Boston Dynamics has been building legged robots longer than any other company on this list. The original Atlas (hydraulic, launched 2013) became the defining image of humanoid robotics for a decade — capable of backflips, parkour, and balance recovery, but never commercially viable at its power and cost profile. In April 2024, Boston Dynamics retired the hydraulic Atlas and unveiled a fully electric version.

Boston Dynamics’ electric Atlas is a technological leap — the electric actuation system allows joint angles beyond the human range of motion, enabling manipulation postures that biomechanically constrained human workers cannot sustain. Hyundai has stated it plans to deploy Atlas in its own manufacturing facilities as the first commercial use case.

Boston Dynamics benefits from the deepest robotics engineering heritage of the group (acquired by Hyundai from SoftBank, which acquired it from Google, which acquired it from its MIT/DARPA roots). The challenge is commercialization: Boston Dynamics has struggled historically to convert research excellence into volume products at scale. Spot (the quadruped) broke that pattern — it is a commercially available product with a published price and active industrial deployments. Whether Atlas can follow Spot’s path is the key question.

Head-to-head comparison

What to watch in H2 2026

1. Tesla Optimus external sales — Does Tesla announce a commercial price and begin shipping Optimus units outside its own factories?
2. Figure AI BMW scale-up — Does the BMW Spartanburg pilot expand to additional assembly stations or additional BMW plants?
3. Agility RoboFab utilization — How many Digit units does Amazon actually deploy beyond Sequoia?
4. 1X NEO home trials — Does 1X publish results from its NEO home-assistant trials, and what tasks can NEO reliably complete?
5. Boston Dynamics Atlas Hyundai deployment — Does Hyundai publish performance data from Atlas in its manufacturing facilities?
6. Humanoid-specific regulation — Will OSHA, ISO, or the EU AI Act impose certification requirements for humanoid robots working alongside humans?

The humanoid race is no longer just about who can make a robot walk. It is about who can make a robot work — reliably, economically, and safely — in the real environments where labor is needed.

Sources

• Figure AI raises $675M Series B — Figure AI press release ↗
• Agility Robotics Digit deployment at Amazon Sequoia facility — Agility Robotics ↗
• Tesla Optimus production ramp — Musk earnings call statements Q1 2026 ↗
• 1X Technologies NEO humanoid — 1X Technologies ↗
• Boston Dynamics Atlas electric reveal — Boston Dynamics ↗
• Hyundai acquires Boston Dynamics — Boston Dynamics press room ↗

Tags