Signals Inbox·July 28, 2026·Humanoid Robotics

Can Agility beat Tesla in the humanoid robot race?

Agility is ahead where humanoid robots are already doing paid industrial work. Tesla is still the more dangerous long-term rival, because once Optimus is reliable, its manufacturing and AI advantages could erase that lead very quickly.

Q1Why is Agility versus Tesla a real humanoid robot race now?

Agility Robotics and Tesla are now in a genuine humanoid robot race because Agility has paying industrial deployments while Tesla has started building the production system meant to overwhelm smaller rivals.

A year or two ago, the comparison was mostly theoretical. Digit was being shaped around warehouse work, while Optimus was still judged through demonstrations and Elon Musk’s promises. Now the gap between the two strategies is much clearer.

Agility has commercial agreements with GXO, Schaeffler, Toyota Motor Manufacturing Canada and Mercado Libre. It is also moving toward a public listing intended to fund a much larger production ramp.

Tesla is approaching the race from the other end. Optimus still lacks a publicly named external customer, but Tesla says its first large production line is being installed in California, with a much larger second-generation line planned in Texas.

The race is easy to describe: Agility is trying to build a customer lead before Tesla turns manufacturing scale into a working robot.

Q2What would it mean for Agility to beat Tesla in humanoid robots?

Agility would beat Tesla by becoming the preferred industrial humanoid platform before Optimus reaches comparable reliability, cost and customer adoption.

Social-media attention tells us little about that outcome. The useful measures are paid deployments, productive hours, repeat orders, safety approval, customer expansion and the cost of completing real work.

The time horizon changes the answer. Agility can lead the market today and still lose the larger race later. Tesla can remain behind in customer deployments while building the stronger long-term production position.

We use a fairly narrow definition: Agility wins if Digit builds a durable lead in factories, warehouses and distribution centers before Tesla can offer an Optimus that customers trust at a lower cost.

What winning the humanoid robot race means

Q3Is Agility already ahead of Tesla in humanoid robots today?

Agility is ahead of Tesla in commercial humanoid robots today. Tesla is better placed to dominate later if mass production actually begins.

Agility’s latest transaction materials say Digit has accumulated more than 65,000 operating hours across commitments involving nine customer facilities. The company also names four large enterprises where Digit is commercially deployed.

Tesla gives investors far less operational detail about Optimus. Its latest official update describes progress “ahead of mass production” and shows the California and Texas robot factories as being under construction. Tesla has yet to disclose a paying external customer, comparable operating hours or customer throughput.

So Agility leads on evidence we can inspect. Tesla’s advantage sits elsewhere: cash, AI infrastructure, supply-chain leverage and experience producing hundreds of thousands of complex machines every quarter.

Current commercial evidence for Digit and Optimus

Q4Has Agility’s Digit actually moved beyond pilot mode?

Digit has moved beyond an ordinary pilot because GXO has used the humanoid robot in day-to-day operations under a multi-year commercial agreement.

The strongest proof is repetition. Agility says Digit has moved more than 100,000 totes at GXO’s Flowery Branch facility, transferring containers from autonomous mobile robots to a conveyor and handling changes when the downstream flow becomes blocked.

A six-figure task count is more useful than a polished demonstration. The robot had to repeat the same physical cycle while dealing with shifting tote positions, warehouse traffic and production interruptions. GXO also moved from an earlier test into a Robot-as-a-Service contract, which is real commitment, not a week-long photo opportunity.

There is still a big hole in the data. Agility has not revealed how many robots produced those movements, the average cycle time, uptime, human intervention rate or savings per tote. Digit is doing real work. Large-fleet profitability is not proven yet.

Q5Is Tesla Optimus doing useful work at scale today?

Tesla Optimus still has no publicly verified record of useful work at commercial scale.

Tesla has shown Optimus sorting objects, walking through facilities and performing manipulation tasks. The demonstrations show technical progress, especially compared with the project’s early years, but Tesla has not released the operating detail needed for a direct comparison with Digit.

The company says Optimus is progressing ahead of mass production and that first-generation production lines are being installed in anticipation of volume output. That puts the robot somewhere between advanced development and industrial production.

Internal factory work could be further along than Tesla has disclosed. But a robot tested inside Tesla gets a friendly environment, immediate engineering support and workflows the company can modify. An outside customer brings procurement rules, insurance, safety teams, integration work and the option to cancel.

For now, Optimus is a high-potential product without a visible commercial track record.

Q6Which humanoid robot is closer to a profitable first job?

Agility’s Digit is closer to a profitable first job because it was designed around repetitive material handling that companies already pay people and automation vendors to perform.

Digit can carry 35 pounds and run for up to four hours before charging. Its early work includes moving totes between mobile robots and conveyors, stacking containers and linking pieces of warehouse automation that were never designed to work together.

These jobs look narrow. That is partly why they can work. The object is known, the route is controlled and the task happens thousands of times. A customer can compare the robot’s cost with overtime, temporary labor, injuries, turnover and lost throughput.

Optimus aims much wider. Tesla wants one robot that can eventually handle factory work, service jobs and domestic tasks. A broader machine could address a far larger market, but it also needs better hands, richer perception and stronger generalization before a first deployment can be trusted.

Agility chose the easier commercial entry point, and right now that choice gives Digit the advantage.

Q7Is Agility’s warehouse focus an advantage or a trap?

Agility’s warehouse focus is helping Digit win customers now, but it could cap the company’s lead if Digit never moves far beyond tote handling.

Warehouses give Agility a practical place to learn. Floors are mostly flat, containers are standardized and workflows are mapped. The company can improve navigation, grasping, charging and exception handling without solving every problem in general robotics.

The sales pitch is simple too. GXO or Mercado Libre does not need to believe in a household robot. The buyer only needs to see a bottleneck that Digit can remove inside an existing building.

The risk is old-fashioned automation. Conveyors, robotic arms, autonomous mobile robots and automated storage systems can often perform one task more cheaply than a two-legged machine. Digit earns its extra complexity only where a facility cannot justify major retrofits, or where one robot can switch between several jobs.

The warehouse strategy buys Agility time and operating experience. Keeping the lead means handling a wider mix of containers, routes and factory tasks without turning every deployment into a custom engineering project.

Q8Can Tesla’s manufacturing machine erase Agility’s lead?

Yes. Tesla can erase Agility’s current lead once Optimus is reliable enough to justify the enormous factories being planned around it.

Tesla produced 451,758 vehicles in its latest reported quarter. That does not prove it can build humanoid robots, but it does show a level of supplier management, factory engineering and quality control that Agility has never operated at.

Tesla says the first Fremont Optimus line is designed for one million robots annually. The later Texas design targets long-term capacity of ten million. Agility’s RoboFab has a stated peak capacity of 10,000 units per year.

Those figures are plans, not output. Tesla lists both robot sites as under construction, and a production line cannot rescue a robot whose hands, actuators or software keep changing. Still, the scale gap is brutal. Even 10% utilization of Tesla’s smaller line would equal ten RoboFabs at their stated ceiling.

Manufacturing position of Agility and Tesla

Q9Can Agility manufacture enough Digit robots to stay ahead?

Agility now has a credible route to thousands of Digit robots, but it has not shown that RoboFab can deliver them quickly or profitably.

RoboFab was built with modular workcells and a peak capacity of 10,000 robots per year. Agility says roughly 75% of Digit’s parts are sourced in the United States and that it owns several of the robot’s highest-value hardware systems. That should reduce some supplier risk and give the company more control over design changes.

Demand is becoming easier to see. Agility reports more than $300 million in multi-year Digit v5 orders, subject to contractual milestones, and a pipeline of more than 30 customers. “Subject to milestones” is doing a lot of work there: announced orders can shrink, move or disappear before revenue is recorded.

Funding is also moving forward. Agility and Churchill Capital have submitted a draft registration statement for their proposed combination. The transaction could provide more than $620 million in gross proceeds, assuming limited shareholder redemptions and a successful close.

Agility has assembled the pieces for a ramp. The missing proof is factory output: completed units, production yield, cost per robot and the speed at which customer orders become working fleets.

Q10Does Agility’s real-world robot data beat Tesla’s AI advantage?

Agility has better task-specific robot data today. Tesla owns the stronger AI and compute infrastructure.

Every Digit shift can produce useful information about grasp failures, balance, tote placement, lighting, congestion and recovery from interruptions. Agility trains skills through demonstrations, teleoperation, reinforcement learning and simulation, then tests them in customer environments.

The company has added a 60,000-square-foot Fremont hub where it plans to hire nearly 200 people across AI, software and field operations. It is also working with NVIDIA hardware and simulation tools to run more demanding perception and manipulation models on the robot.

Tesla operates at another level of compute. Cortex 2 has started training workloads with planned capacity above 130,000 H100-equivalent GPUs, and Tesla is developing Dojo 3 and its own inference chips. Its vehicle program has already forced the company to solve large-scale vision, planning, training and deployment problems.

Road data cannot teach a robot hand to grip a flexible object or recover after bad contact. Agility’s smaller dataset is closer to the warehouse problems Digit needs to solve. Tesla becomes much more dangerous once thousands of Optimus units begin producing manipulation data of their own.

Q11Who is ahead on humanoid robot safety, Agility or Tesla?

Agility has the stronger public safety record for humanoid robots working inside industrial facilities.

Digit passed a site-specific evaluation by an OSHA-recognized Nationally Recognized Testing Laboratory at a live fulfillment site. The assessment covered mechanical, electrical and interaction hazards, the sort of issues that can stop an enterprise deployment before it starts.

Agility has also added a safety PLC, emergency stops, a controlled-stop function and other industrial safety features. Its engineers are contributing to emerging ISO and ANSI work for dynamically stable industrial mobile robots.

Digit v5 is designed for cooperative safety, allowing robots and people to share dynamic spaces without fixed barriers. That promise still needs commercial proof, and Agility says current deployments rely on separation.

Tesla may have extensive internal Optimus safety work, but it has not published a comparable third-party site evaluation or customer approval process. On the public evidence, Agility is comfortably ahead.

Q12Can Agility beat Tesla on humanoid robot costs before Optimus scales?

Agility could win the early cost battle because Digit already has customers, although nobody outside those contracts can verify its true cost per productive hour.

The Robot-as-a-Service model lowers the buyer’s risk. Customers can pay for a managed service that includes the robot, software and support rather than buying an expensive machine before its performance is proven.

Digit also targets jobs with visible labor costs. A warehouse operator can compare the service fee with staffing gaps, overtime, temporary workers, injury exposure and missed throughput. That makes the purchase easier to defend internally.

The missing numbers block a firm economic verdict. Agility does not publish service pricing, gross margin, maintenance hours, remote assistance or the cost of each completed task. The 100,000-tote deployment confirms repeated use; it says almost nothing about profit.

Tesla’s price advantage would come later through production scale and vertical integration. Agility’s opening is to prove savings while Optimus is still unavailable to outside buyers.

Q13Does Tesla’s money make Agility’s defeat inevitable?

Tesla’s financial strength gives Optimus more chances to succeed. It does not guarantee that Tesla will build the better industrial robot.

Tesla ended its latest reported quarter with about $44.7 billion in cash, cash equivalents and short-term investments. It also generated roughly $3.9 billion in operating cash flow during that quarter. Agility is trying to secure a little over $620 million through a transaction that still requires regulatory review and shareholder approval.

That difference changes how each company can handle mistakes. Tesla can fund several hand designs, custom chips, new production equipment and years of AI training. Agility has less room for a delayed product generation or a factory ramp that misses its targets.

Agility does get one benefit from being smaller: focus. Digit is the company’s central product, while Tesla is simultaneously scaling vehicles, energy storage, robotaxis, semiconductors, AI infrastructure and Optimus. Money reduces technical risk. It does not create more management attention.

Tesla can afford a longer race. Agility has to use its current commercial lead before the funding gap starts shaping product quality and price.

Q14Could Agility win industrial humanoids while Tesla wins the broader robot market?

Yes. Agility can become the leading industrial humanoid supplier even if Tesla eventually sells far more robots overall.

Industrial robotics already supports several winners. Fixed arms dominate repeatable manufacturing cells, autonomous mobile robots move goods across floors, and specialized systems handle storage, picking or inspection. Humanoids are unlikely to erase that variety.

Digit fits jobs where facilities were designed for people and fixed automation leaves awkward gaps. Agility Arc also connects robots with warehouse and manufacturing systems, giving customers a fleet-management layer rather than a standalone machine.

Tesla’s ambition reaches further. Optimus is meant to start in factories and eventually enter service work and homes. Success there could create a market much larger than Agility’s initial logistics and manufacturing base.

Where Agility and Tesla are better placed

Q15What must Agility do before Tesla Optimus reaches outside customers?

Agility must turn Digit’s early deployments into large, sticky customer fleets before Tesla offers Optimus to the same buyers.

The immediate problem is converting conditional orders into delivered robots and recognized revenue. A $300 million order book sounds impressive, but customers can delay expansion when milestones, safety requirements or expected economics are missed.

Digit also needs a wider set of skills. Machine tending, kitting, line-side delivery and several container types would give customers more reasons to keep the same fleet when a cheaper rival appears.

The harder part is proving the economics publicly. Uptime, interventions per shift, maintenance, cycle time and cost per task would carry more weight than another demonstration. Buyers could finally compare Digit with conventional automation and future Optimus offers.

Arc, employee training, safety documentation and workflow data can then make Digit harder to replace across multiple sites. The customer list gets Agility into the building. Expansion across tasks and sites is what might keep Tesla out.

Q16What would prove Tesla has taken the lead from Agility?

Tesla takes the humanoid robot lead when Optimus combines repeatable work, outside customers and meaningful production volume.

A finished-looking robot or another factory video would fall short. We would need several months of operating data from Tesla facilities, including task throughput, human interventions and uptime. External contracts would then show whether independent buyers accept the performance, safety process and price.

Production is the final test. Tesla’s million-unit line becomes relevant when it ships consistent robots, not when equipment is installed. Even tens of thousands of reliable units would put enormous pressure on Agility because Tesla could gather more data, negotiate lower component prices and support more customers.

Until those three conditions appear together, Tesla remains the scale favourite rather than the commercial leader.

Q17Can Agility beat Tesla in the humanoid robot race?

Yes. Agility can beat Tesla in industrial humanoid robots, but Tesla is still more likely to lead the overall market if Optimus survives the jump from prototype to mass production.

Agility owns the clearest commercial advantage today. Digit works for named outside customers, has crossed meaningful operating and throughput milestones, and has passed a third-party safety evaluation at a customer site. The new Fremont AI hub, conditional Digit v5 orders and proposed public transaction show that Agility is preparing for a larger push rather than protecting a small pilot business.

The weaknesses are just as clear. We have little visibility into Digit’s productivity, service costs, factory output or customer savings. Agility’s announced production ceiling is also tiny beside Tesla’s planned capacity.

Tesla has the opposite problem: Optimus lacks a comparable external deployment record, while Tesla already has the money, compute, factories and supply-chain experience needed to scale very fast once the product is ready.

Agility is winning the industrial deployment phase. It can keep that lead by turning a handful of customer sites into large fleets with several profitable tasks. If Digit remains expensive and narrowly useful when Optimus reaches outside customers, Tesla will probably pass it.