Figure 03 vs Tesla Optimus Comparison Tracker (2026)
In our humanoid robotics market deck, you will find everything you need to understand the market
Summary
Tesla Optimus and Figure 03 are the two humanoid robots investors and operators ask about most in 2026. This tracker answers one simple question: which one is actually further ahead today?
To answer it, we built a 15-question head-to-head across three categories. Each question was scored from 0 to 100 using only tier-one public evidence. Categories were then weighted by how much each one reflects today's commercial reality rather than future ambition.
The short answer is that Figure is ahead today by a meaningful margin. Figure scored 78.9 out of 100, and Tesla scored 45.1. Figure leads every category, but the overall gap is driven mostly by one of them.
Real-world traction is where the gap is widest. Figure scored 79 there and Tesla scored 32. That 47-point difference comes from BMW's own published numbers, not from Figure marketing. On pure robot performance specs, the gap narrows to about 26 points. The decisive issue today is deployment proof, not engineering capability.
Figure has a named paying customer already doing productive work. Figure 02 helped BMW build over 30,000 X3 vehicles during a 10-month deployment at Plant Spartanburg. BMW is now expanding Physical AI to Plant Leipzig with Figure 03. Tesla has zero announced external customers as of April 2026. On the Q4 2025 earnings call, Musk said Optimus "is not in usage in our factories in a material way."
Several of Tesla's most visible Optimus demos have also been reported as teleoperated. Bloomberg flagged this first at the October 2024 Cybercab event. Electrek, TechCrunch and The Verge have documented similar patterns since. The December 2025 Miami demo is the latest example, where a robot tipped over with hand motions that looked like a remote operator removing a headset.
Tesla's long-term case is still serious. Fremont line conversion, the AI5 chip, Cortex 2.0 compute, and over 41 billion dollars in cash give Tesla scale options no other humanoid maker has. But "best positioned to win later" and "ahead today" are different judgments. By the current-evidence standard, Figure clearly leads.
This market map, featured in our humanoid robotics market deck, highlights top companies and startups in the humanoid robotics market
Tesla vs Figure Comparison Tracker
This tracker compares Tesla Optimus and Figure 03 across 15 evidence-based questions, grouped into three categories: robot performance, real-world traction, and learning and trajectory. Each question was scored from 0 to 100 using only tier-one public evidence as of April 2026. That includes company press releases, earnings call transcripts, named executive statements, and reporting from Bloomberg, The Verge, Electrek, TechCrunch and The Information that traces back to primary sources.
Figure's overall weighted score came to 78.9 and Tesla's to 45.1. Figure leads every category. The decisive gap sits in real-world traction, where Figure has a paying customer in BMW and Tesla has zero announced external customers. If you want the broader market context on how this head-to-head fits within the humanoid robotics sector, we go deeper in our humanoid robotics market deck.
| Question | Category | Figure evidence | Figure score | Tesla evidence | Tesla score | Winner |
|---|---|---|---|---|---|---|
| How much weight can it handle credibly? | Robot performance | Figure 03 carries 20 kg with 2x faster actuators. It moved over 90,000 sheet-metal parts across 1,250 hours at BMW under real production load. | 78 | Optimus specs 20 kg carry and 68 kg deadlift, with footage of battery trays and stair climbing. No external benchmark or cycle-time data has been published. | 62 | Figure |
| How long can it work on a full battery? | Robot performance | Figure 03 is rated at about 5 hours per charge with 2 kW inductive charging. BotQ units now run 24/7 via charging mats, and BMW ran 10-hour shifts daily. | 80 | Optimus has a 2.3 kWh battery rated for 2 to 4 hours of dynamic walking. No continuous runtime proof under production load has been shown externally. | 60 | Figure |
| How proven is its durability in sustained use? | Robot performance | Figure 02 ran 1,250 hours on an active BMW line and came back visibly scuffed. Figure published the wear data and rolled the lessons into Figure 03. | 82 | The Information and LatePost Auto reported a mid-2025 production pause. The causes were overheating joints, short transmissions, and weak hands. Musk has acknowledged upper limbs are still unsolved. | 40 | Figure |
| How safely can it work around humans? | Robot performance | Figure 03 uses soft textiles, multi-density foam, a UN38.3-certified battery, and 3 g tactile fingertips. It ran near BMW line workers for 11 months with no reported incidents. | 74 | Optimus lists quiet operation and force sensors. In December 2025 an Optimus tipped over in Miami with hand motions suggesting a teleoperator removing a VR headset. | 50 | Figure |
| What tasks does it perform most credibly today? | Robot performance | Loaded sheet metal into welding fixtures on an active BMW X3 line for 10 months. Helix 02 ran a 4-minute, 61-action dishwasher task with no resets and no human intervention. | 80 | Tesla shows autonomous 4680 cell sorting, AI-driven kung fu, and jogging. We Robot 2024 and Autonomy Visualized 2025 demos were reported by multiple outlets as teleoperated. | 52 | Figure |
| Has it been deployed by a third-party customer? | Real-world traction | BMW Plant Spartanburg confirmed a 10-month deployment producing 30,000+ X3 vehicles. BMW is now expanding Physical AI to Leipzig in Germany with Figure 03. | 88 | No external customer deployment has been confirmed. Musk stated on Q4 2025 that Optimus "is not in usage in our factories in a material way." | 25 | Figure |
| Is there evidence of sustained use over time? | Real-world traction | BMW and Figure jointly report 10-hour shifts for nearly a year, 1,250 hours, 90,000+ parts loaded, and about 1.2 million robot steps. | 86 | Tesla has disclosed no cycle times, uptime, or MTBF figures. The Q4 2025 call said units are "primarily for learning, not productive tasks." | 30 | Figure |
| Has it done useful work in a real setting? | Real-world traction | Figure 02 contributed directly to production of 30,000+ X3 vehicles. BMW's own press statement calls the value "measurable" under real-world conditions. | 88 | Musk said on Q4 2025 that "we wouldn't expect any significant Optimus production volume until probably end of this year." Current units exist to learn. | 25 | Figure |
| Is there credible customer demand or signed interest? | Real-world traction | BMW signed January 2024 and is now expanding to Leipzig with F.03. A second U.S. customer is active, reported as UPS by Bloomberg and Fortune. | 78 | No external commercial customer has been announced. Musk said external sales "probably" start late 2026 or 2027. All orders remain internal. | 35 | Figure |
| How many robots are credibly active in real deployments? | Real-world traction | Small fleet at BMW plus units at a second paying customer. Sunnyvale is running 24/7. The external commercial count is still modest, measured in tens of robots. | 55 | Third-party tracking cites "over 1,000" Optimus units at Fremont and Giga Texas. Musk himself says they are not doing productive work. | 45 | Figure |
| How fast is visible performance improving? | Learning and trajectory | Figure 01 to 02 to 03 in two years, with a claimed 90% component cost cut, 2x actuator speed, 3 g tactile fingertips, and full Helix AI integration after ending the OpenAI partnership. | 82 | Gen 2 to Gen 3 took longer than prior Musk timelines and required replacing the Optimus program lead. A new 22-DoF, 50-actuator hand was revealed in February 2026. | 58 | Figure |
| How fast is autonomy improving? | Learning and trajectory | Helix 02 in January 2026 added System 0/1/2 full-body autonomy. A 10M-parameter network replaced about 109,000 lines of hand-engineered C++ balance code. | 82 | Tesla showed AI-driven kung fu and autonomous sorting. But Bloomberg, TechCrunch, The Verge and Electrek have repeatedly flagged teleoperation at the most visible demos. | 50 | Figure |
| How fast is task variety expanding? | Learning and trajectory | Helix runs sheet metal loading, parcel sorting, laundry folding, and dishwasher loading on one architecture. Changes between use cases are data-only. | 80 | Tesla has shown factory sorting, kitting, jogging, and staged household clips. Most household demos were reported as teleoperated rather than autonomous. | 50 | Figure |
| How strong is the evidence of learning from data? | Learning and trajectory | Figure's first use case took 12 months to deploy. The second took 30 days. That curve is the clearest public proxy for data compounding in humanoid robotics. | 78 | Tesla's data flywheel rests on thousands of internal units, FSD-v15 architecture sharing, and a forthcoming Cortex 2.0 compute stack. Still mostly anticipatory. | 55 | Figure |
| What is the most realistic timeline for scaled deployment? | Learning and trajectory | BotQ has 12,000-unit annual capacity with a 100,000-unit four-year goal. Figure is already in live production at a paying customer, which is the prerequisite. | 72 | Musk confirmed V3 production starts summer 2026, with high volume in summer 2027. Fremont S/X lines are being converted to Optimus at significant capex. | 62 | Figure |
This chart, featured in our humanoid robotics market deck, compares the main business model options for humanoid robot manufacturers
Insights
The study behind this tracker reviewed every major public statement from Elon Musk and Brett Adcock since 2021. It also covered every BMW, Tesla and Figure press release on humanoid deployment, every relevant earnings call transcript, and every tier-one news report that could be traced back to a primary source. The goal was not to compare ambition, since both companies have plenty of it. The goal was to compare what each one has actually proven in public as of April 2026.
Some of the most useful patterns only become visible once Tesla Optimus and Figure 03 evidence is placed in the same analytical frame. Each insight below is written to stand on its own, so it can be understood without the rest of the article as context.
- In the Tesla Optimus vs Figure 03 comparison, the most important asymmetry is the type of evidence each company publishes. Figure publishes customer-corroborated operational numbers like 30,000+ BMW vehicles produced and 1,250 hours of runtime. Tesla publishes internal unit counts, spec targets, and forward-dated milestones.
- Tesla Optimus has a competitive hardware spec on paper. The gap to Figure 03 narrows to about 26 points in the robot performance category of this tracker. The decisive gap in the Tesla vs Figure comparison is real-world traction, at 47 points, not engineering capability.
- Elon Musk himself validated the core thesis that Tesla Optimus is behind Figure 03 in commercial deployment. On Tesla's Q4 2025 earnings call, he said Optimus "is not in usage in our factories in a material way." That single sentence is the clearest concession in the entire public Tesla humanoid record.
- BMW's March 2026 expansion of Figure's humanoid program from Plant Spartanburg to Plant Leipzig is customer-initiated. That is much harder to fake than an investor-facing demo. Customers pulling a vendor's robot into a second geography is the kind of signal that historically precedes broader category adoption.
- BMW chose Figure in January 2024, at a time when Tesla Optimus was already equally famous in humanoid robotics. The "brand pulls procurement" thesis does not hold in enterprise humanoids. Technical maturity, not CEO attention, drove the BMW decision.
- Figure's first industrial use case at BMW took 12 months to deploy. The second use case at a separate paying customer took 30 days. That 12-to-30 curve is the clearest publicly visible signal of AI-driven deployment economics in humanoid robotics, and it has no Tesla equivalent yet.
- Figure's Helix 02 AI system, shipped in January 2026, runs a 10-million-parameter neural network. It replaced roughly 109,000 lines of hand-engineered C++ balance code. This is the clearest public signal of genuine embodied AI progress in the humanoid sector, not just a cosmetic demo upgrade.
- Tesla's Optimus demos in 2024 and 2025 show a recurring teleoperation pattern. Bloomberg, The Verge, TechCrunch and Electrek separately flagged teleoperation at We Robot 2024 and Autonomy Visualized 2025. The market has not yet fully re-priced this credibility loss.
- At Tesla's December 2025 Autonomy Visualized event in Miami, an Optimus robot tipped over with hand motions that closely matched a VR-headset removal. That specific motion is hard to explain under an autonomous-operation hypothesis. Gene Munster and Robert Scoble were among the eyewitnesses who flagged it publicly.
- The market still rewards Tesla's future humanoid manufacturing capacity more than Figure's present customer output. In humanoid robotics, present output matters more, because the scarce input is real-work data, not assembly-line throughput. That framing is structurally backwards for this sector.
- Tesla's Optimus program lead Milan Kovac left in June 2025, in the middle of the production ramp. Ashok Elluswamy, head of Autopilot AI, took over. Mid-ramp leadership transitions rarely accelerate complex hardware programs.
- Tesla paused Optimus production in mid-2025 over overheating joint motors, short transmission life, and weak hand performance, per The Information and LatePost Auto. Musk has publicly acknowledged that upper limb reliability remains unsolved. The issues surfaced in the press before Tesla confirmed them.
- Agility Robotics has moved over 100,000 totes at GXO under a multi-year Robot-as-a-Service agreement. Toyota and Mercado Libre were added as customers in 2025. The title of "first commercially productive humanoid robot" arguably belongs to Agility, not Figure or Tesla.
- Every major automaker except Tesla is partnering with a specialist humanoid robotics company. Hyundai backs Boston Dynamics, Mercedes backs Apptronik, BMW backs Figure, and Toyota backs Agility. Tesla is the only large carmaker building its own humanoid in-house rather than partnering.
- Chinese humanoid robot volume could eventually dwarf both Tesla Optimus and Figure 03. AgiBot alone reported over 1,000 humanoids built in 2024. Chinese government targets point to 59 million humanoids in domestic deployment by 2050, which is an order of magnitude larger than any Western humanoid target.
- Figure's September 2025 Series C closed at a 39 billion dollar valuation. It attracted Nvidia, Microsoft, Bezos Expeditions, Intel Capital, Salesforce, LG, Qualcomm and T-Mobile. Tesla is vertically integrating Optimus instead of building a partner set like this, which leaves it more exposed to single-point-of-failure risk.
This chart, featured in our humanoid robotics market deck, shows how Agility Robotics is capturing share in humanoid robotics
Why is Figure executing better than Tesla today?
Figure is executing better than Tesla today because it has converted its robot into real commercial work at BMW. Tesla's Optimus, by Musk's own admission, is not yet doing useful factory work.
Figure 02 ran Monday-to-Friday shifts at BMW Plant Spartanburg for 10 months. It helped produce more than 30,000 X3 vehicles and loaded over 90,000 sheet-metal parts. Those numbers come from BMW's own press statements, which makes them much harder to discount than typical vendor claims.
Tesla's clearest acknowledgment of the gap comes from Musk himself. On the January 2026 Q4 earnings call, he said Optimus "is not in usage in our factories in a material way." That quote is sourced directly from Tesla's CEO on a legally sensitive call. It is the kind of statement no external analyst can match.
Figure also shipped Helix 02 in January 2026. It completed a 4-minute, 61-action dishwasher task end-to-end with no resets. Tesla's most visible 2025 demos were repeatedly flagged as teleoperated by Bloomberg, The Verge and Electrek. The contrast between credibly autonomous and credibly teleoperated is doing most of the work in how each company's 2026 credibility is being priced.
Tesla's 2025 program stability also lagged. Optimus program lead Milan Kovac left in June 2025. The program reportedly paused over hand and arm reliability issues. Those setbacks compound rather than resolve in a single quarter, which is why the gap widened across 2025 instead of closing.
By the way, this execution gap is where the investor narrative is currently most mispriced. We go deeper on it in our humanoid robotics market deck.
Do Tesla and Figure target the same initial market?
Tesla and Figure do not fully target the same initial market today. Figure is selling to third-party industrial and logistics customers, while Tesla's current "customer" is Tesla itself.
Figure's first customer is BMW, with a commercial agreement signed in January 2024. Its second customer is reportedly UPS, per Bloomberg and Fortune reporting. Both are Fortune 500 operators under formal procurement. They come with real ROI thresholds and integration requirements that a demo alone cannot clear.
Tesla's Optimus units operate inside Fremont and Giga Texas. They do 4680 battery cell sorting, parts handling, and kitting. Musk said on the Q4 2025 call that external commercial sale is targeted for late 2026 at the earliest, with consumer sales guided to 2027. That puts Tesla's external market roughly 18 months behind Figure's on a simple calendar basis.
The two strategies do converge in 2026 to 2027. Figure adds home deployments with Figure 03. Tesla begins external industrial sales. Until that convergence, their demand pools are structurally different. This matters more than it looks, because early humanoid procurement choices are locking in integrations that will be hard to reverse once plants are certified.
There are real overlap signals too. Both target 20,000 to 30,000 dollar unit prices at scale. Both built their own AI stacks. Both are drawing from the same VC, strategic investor, and talent pools. The competitive overlap is genuine, it is just not yet active.
What advantages does Tesla have that Figure lacks?
Tesla's decisive advantages over Figure are manufacturing scale, capital base, and in-house AI compute. Several of its other commonly cited advantages are less transferable to humanoid robotics than the market assumes.
Tesla ended Q3 2025 with over 41 billion dollars in cash and investments. Figure has raised roughly 1.9 billion in total. That gap lets Tesla absorb multi-year slippage without financial stress. It is a genuine structural edge in a sector where cost-of-failure matters more than cost-of-capital.
Tesla is also converting Fremont Model S/X lines to Optimus production. Giga Texas is planned for 10 million units per year at full buildout. No other humanoid maker has that manufacturing depth. The Fremont conversion alone represents capex on the order of 20 billion dollars.
Tesla's advantages split cleanly between decisive and overstated. It is worth being precise about which is which:
- Genuinely decisive: Fremont line conversion, AI5 chip, Cortex 2.0 compute, vertically integrated actuators and batteries
- Overstated: brand as procurement lever, since BMW chose Figure in January 2024 when Optimus was already equally famous
- Latent: manufacturing scale matters only once a working, certified Optimus V3 exists at spec, not before
The most useful way to read Tesla's position is that its advantages are real but time-shifted. They kick in after Optimus V3 production begins in summer 2026, not today. That is why Figure's current lead is not yet threatened by them.
This chart, featured in our humanoid robotics market deck, illustrates yearly funding for humanoid robotics startups
Will Figure deploy humanoid robots at scale before Tesla?
Figure is likely to deploy humanoid robots at scale before Tesla. We put the probability in the 60 to 75 percent range, because Figure is already in live external commercial production while Tesla's Gen 3 Optimus only starts in summer 2026.
Figure has a paying customer that is already expanding geographically. BMW moved from Spartanburg to Leipzig in March 2026. BotQ has a 12,000-unit annual design capacity and a 50,000-unit target for 2026. That is a real head start, not a narrative one. Customer-initiated expansion is the hardest signal in enterprise hardware to fake.
Tesla has zero announced external deployments. Gen 3 production starts in summer 2026 per Musk at the Abundance Summit. High-volume production is guided to summer 2027. On a simple timeline basis, Figure is roughly 12 to 18 months ahead. That lead is what underpins the 60 percent lower bound.
The probability is capped at 75 percent because Tesla's manufacturing ramp, once it turns on, is designed for automotive volumes. Fremont alone targets 1 million units per year at full conversion. That dwarfs BotQ's first-gen capacity. Scale curves at Tesla volumes have historically beaten startup curves, which is why the probability does not push into the 80s.
One thing matters in reading this probability: the definition of "scale." It means thousands of robots doing productive work at third-party customers, not internal fleets. By that standard, Figure starts from a positive base and Tesla starts from zero. That is why the probability sits above 60 percent rather than below it.
By the way, we go deeper on these timelines and cost curves in our humanoid robotics market deck.
Which milestones would most change Tesla vs Figure odds?
Three milestones would move the Tesla vs Figure odds the most. Tesla hitting V3 production on schedule, Figure signing a third named commercial customer, and either company publishing real operating metrics for the first time.
Below are the most odds-moving milestones to watch over the next 12 to 24 months:
- Tesla V3 production actually starting in summer 2026, sustained above 1,000 units per month
- Figure signing a third or fourth named commercial customer with publicly verified operational data
- Tesla's first external customer announcement, currently targeted for late 2026
- Figure 03 reaching alpha home testers with published reliability and safety numbers
- Either company publishing cost per task, uptime, or mean time between failures for the first time
- Safety certification events such as UL or OSHA NRTL, following Agility Robotics' 2025 precedent
Missing V3 production by more than two quarters would be Tesla's third major humanoid slip, after 2024 and 2025. At some point, slip history becomes a probability update the market cannot ignore. The third consecutive slip is historically where that repricing tends to happen in hardware sectors.
Could Tesla fail to lead humanoid robotics long-term?
Tesla could fail to lead humanoid robotics long-term. The probability sits in the 30 to 45 percent range, because the current gap plus Tesla's 2025 execution record makes "never leads" a real possibility despite its structural advantages.
Tesla already missed its own 2025 production target for Optimus. Musk committed to 5,000 to 10,000 units on the Q4 2024 call. Actual 2025 output was in the low hundreds to low thousands, per The Information and LatePost Auto. Missing a self-set production target by more than 80 percent is the kind of data point that compounds in enterprise procurement diligence, not just in investor narratives.
Enterprise procurement in humanoids is also locking in now. BMW chose Figure. Mercedes chose Apptronik. Hyundai backs Boston Dynamics. Toyota backs Agility. Tesla has zero OEM customer pairing, and switching costs grow once a humanoid vendor is integrated into factory MES and PLM systems.
The "data flywheel" argument only pays off with productive external work. Musk said on Q4 2025 that Optimus is "not doing useful work" yet. Meanwhile Figure and Agility are already collecting real external task data. The data that actually trains useful humanoid AI is real-work data, and Tesla currently has materially less of it than either leader.
The 30 to 45 percent range reflects the view that Tesla's capital and manufacturing advantages are genuine, but they still need to land on time to convert. If Optimus V3 slips another year, the probability moves toward the upper end of the range quickly.
By the way, we go deeper on Tesla's structural risk factors in our humanoid robotics market deck.
This chart, featured in our humanoid robotics market deck, shows how warehouse automation has driven growth in the humanoid robotics market over time
Will Tesla dwarf Figure in humanoid robots long-term?
Tesla dwarfing Figure long-term is plausible but not probable. We put the probability in the 30 to 45 percent range, because Tesla's manufacturing and capital scale make it structurally possible but require execution Tesla has not yet shown in humanoids.
"Dwarfing" here means Tesla reaching more than 10 times Figure's deployed units, revenue, or cost position over multi-year timeframes. That is a high bar. It requires several variables to line up at once, not just one breakthrough moment.
Tesla would need to hit summer 2026 V3 production on schedule. It would need to ramp Fremont to five-figure monthly output by end of 2027. And it would need to convert external customers on the announced timetable. Missing any one of those three has historically been enough to stall a hardware ramp, and Tesla has already missed its prior humanoid production targets twice.
Figure would also need to stall for the dwarfing scenario to play out. Figure already claims a 90 percent component cost cut from F.02 to F.03. BotQ is scaling to 50,000 units per year. BMW is expanding to a second country. Each quarter of continued Figure execution makes structural dwarfing meaningfully less likely.
The 30 to 45 percent range reflects a genuine underlying tension. It captures Tesla's long-term manufacturing advantage against Figure's present-day commercial lead. In humanoid robotics, the latter tends to compound faster than the former, which is why the probability sits below 50 percent rather than above it.
Could another company beat both Tesla and Figure?
Another company could beat both Tesla and Figure in humanoid robotics. The probability sits in the 30 to 45 percent range, because the market is still early and Agility Robotics currently has more operational proof than either.
The most credible non-Tesla, non-Figure long-term winners, based on current operational evidence rather than ambition, include:
- Agility Robotics, with 100,000+ totes moved at GXO warehouses and new customers Toyota and Mercado Libre
- Apptronik, deployed at Mercedes-Benz Berlin-Marienfelde with Google and Nvidia as AI partners
- Chinese players AgiBot, Unitree, XPENG and Fourier, whose combined volume may dwarf Western output
- 1X, whose NEO opened consumer preorders with 2026 delivery for the home market
- Boston Dynamics, backed by Hyundai, with a physical capability envelope that exceeds both Optimus and Figure 03
Market leadership may ultimately be decided by aggregate Chinese production. AgiBot alone reported over 1,000 humanoids built in 2024. Chinese government targets point to 59 million humanoids in domestic deployment by 2050. That volume could reorder the category on unit count before Tesla or Figure reach their stated long-term targets.
The Tesla-vs-Figure frame itself may be under-reading Agility in particular. Digit already has operational metrics that neither Optimus nor Figure 03 have published publicly. For investors treating the market as a two-horse race, the bigger risk may be a third horse quietly pulling ahead without the narrative attention.
By the way, we compare every major humanoid player head-to-head in our humanoid robotics market deck.
Could Figure become an acquisition target for Tesla or tech giants?
Figure becoming an acquisition target is unlikely in the near term but plausible in a specific distress scenario. Figure's 39 billion dollar valuation and founder ambition block most would-be acquirers under normal conditions.
Only a handful of companies could absorb a 50 billion dollar plus humanoid acquisition. The viable buyer set is limited to Apple, Microsoft, Google, Amazon, Meta, Nvidia, and Tesla. Several are already Figure investors. That actually makes them less likely to acquire outright, not more, because they already capture upside without the antitrust cost.
Founder incentives also cut sharply against acquisition. Adcock has publicly said Figure could become "the biggest company in the world." He previously built Archer Aviation and took it public rather than selling. Founder-pattern bias is a real factor in M&A probability modeling.
Tesla acquiring Figure is strategically coherent but narratively awkward. It would effectively admit Optimus is behind in its own category. That carries a real cost for Musk specifically, given his public positioning of Optimus as Tesla's "biggest product of all time." The narrative cost is a genuine deal-blocker, not just a pride one.
Acquisition becomes more likely only in a specific distress scenario. A Figure safety incident. A home-deployment setback. A macro pullback forcing down-round conditions. Or a Tesla V3 stumble that forces Tesla to buy its way in. Each quarter of BMW expansion and BotQ throughput growth moves Figure further away from acquirable status, not toward it.
Our methodology to build this tracker
We prioritized tier-one evidence only. That includes company press releases from Figure, Tesla, BMW, Intel Capital, Apptronik, Agility and Mercedes-Benz, earnings call transcripts, on-record statements from named executives including Adcock, Musk, Burzer and Nedeljković, and reporting from Bloomberg, TechCrunch, The Verge, Electrek, The Information, Reuters and The Robot Report that traces back to primary sources. Where claims came from secondary reporting only, we discounted them.
We grouped the 15 questions into three categories. Robot performance covers hardware capability under load. Real-world traction covers external customer deployment and sustained use. Learning and trajectory covers AI progress, hardware iteration speed, and realistic deployment timelines. Each question was scored from 0 to 100 based on strength of public evidence, not on ambition or narrative.
We applied category-level weighting rather than per-question weighting. Robot performance is weighted at 35 percent. Real-world traction is weighted at 40 percent. Learning and trajectory is weighted at 25 percent. Traction got the highest weight because current external proof matters slightly more than technical promise alone in this market.
Category scores are simple averages within each category. Figure's category scores came out to 78.8 for robot performance, 79.0 for real-world traction, and 78.8 for learning and trajectory. Tesla's came out to 52.8, 32.0 and 55.0 respectively. Figure's final weighted score was 78.9. Tesla's was 45.1. The distinction between "ahead today" and "best positioned long term" was kept explicit throughout, because both can be honestly true at the same time. Where demos were credibly reported as teleoperated, we discounted them. Where wear marks, runtime hours, or independent customer statements existed, we weighted them heavily.
Related blog posts
- How strong is fundraising in the humanoid robotics market right now?
- How close is Tesla Optimus to being deployed at scale?
- The full range of business models in the humanoid robotics market
