For a month, this thread has dismantled the mechanics of humanoid deceit layer by layer. First the rigged demonstration — the pre-loaded Tai-Chi, the Atlas piloted via VR headset, the backflip-brochure. Then the word: 'autonomous', brandished five times in a week and cracking at every third of a sentence. Then the scene: a televised gala where the broadcaster turned out to be a shareholder in the robots it was making dance. At each level, the same reflex: to retreat a step towards what seemed more solid. One last refuge remained, the only one believed to be incorruptible — the hard metric, the delivery figure, the number of units out of the factory. This week, the deception entered there too.
Between 27 January and 2 February, three players published, in six days, three mutually exclusive 'number one in the world' rankings. Each crafted its own scoreboard. No longer ask who is number one: ask who holds the calculator, and who pays for it to display that number.
Three crowns for a single head
The first to don the crown is AgiBot. On 27 January, the Chinese manufacturer issued an official press release echoing a ranking by the analytics firm Omdia, which places it first globally in humanoid robot deliveries for 2025: 5 168 units, including 3 588 bipedal humanoids from the A and X series and 1 412 mobile manipulators from the G series. Three days later, on 30 January, AgiBot opened its first European event in Milan, at the Magna Pars venue, arriving wrapped in the same figure: 39% global market share according to Omdia, out of a total of approximately 13 000 units sold in the year. The statistical coronation is not a public relations detail: it is the very sales argument for entering Europe. You do not sell a robot, you sell a rank.
Except the rank is contested — by the very one Omdia had ranked second. On 29 January, Unitree Robotics published its own press release and claimed to have delivered more than 5 500 pure humanoid robots to customers in 2025, excluding wheeled robots. The firm had only attributed 4 200 to it. This is the first time Unitree has published delivery figures — and it did so not to submit to the reference source, but to publicly recuse it. The trade press recorded the break without ambiguity: Caixin headlined 'Unitree Defends Robot Sales as Rival Claims Market Crown', Gasgoo asked 'Annual Champion Changes Hands?'. The company, rather than entering the ranking through the door, dismantled the lock.
The only man without a figure promises the biggest figure
On the other side of the Pacific, on 28 January, the third player played a card of a different nature. During Tesla’s fourth quarter 2025 earnings call, Elon Musk admitted being unable to provide any Optimus production figure for 2025, with actual production remaining far below announced targets. He specified that the few Optimus units present in the factories perform only simple tasks — sorting battery cells, for example — for learning purposes and not production, and that the programme remains in the research and development phase: the robots are not deployed 'in a material way'. No significant volume is expected, he added, before probably the end of 2026, and he gave no quantified production target for 2026.
Yet the same man, in the same call, dropped the biggest figure of the entire week: a targeted capacity of one million Optimus units per year at the Fremont factory. To do this, he sacrifices his prestige cars. 'It’s time to basically bring the Model S and X programs to an end', he declared, adding: 'We expect to wind down S and X production next quarter.' The halt of Model S and Model X is scheduled for the second quarter of 2026, and the Fremont line will be reconvered to robot manufacturing. Zero verifiable units on one side, one million promised on the other — from the same mouth, in the same sentence.
The Davos promise collapses in the spreadsheet
The shift is all the clearer because it contradicts a scene this thread noted the previous week. At Davos, six days earlier, Musk promised 'complex tasks by end of 2026' and a 'public sale by end of 2027'. The 28 January earnings call brings this promise back to earth: no Optimus does useful work today, and no volume will arrive before probably the end of 2026. The press noted it without beating around the bush — Electrek summarises: 'Musk admits no Optimus robots are doing useful work — after claiming otherwise.' The promise of the scene does not survive the passage through the spreadsheet.
The motive no one crosses: there is no consensus count
If one aligns these three claims, a detail jumps to the eye: none rests on the same tally. Omdia advances 13 318 units globally for 2025. The China Machinery Robot Association, for its part, estimates approximately 20 000 units for Chinese deliveries alone for the year — that is, by themselves, more than Omdia’s global total. The classification scopes differ: what is called a 'delivered humanoid robot' does not have the same boundary depending on the office counting. Unitree over-counts by 1 300 units compared to Omdia’s ledger; Tesla does not count itself at all, lacking production. The metric this thread spent a month digging out from under the folklore has, in turn, become the folklore: each manufacturer is now its own statistics office.
This arithmetic disorder is not just a quarrel of accountants. It thrives on a technical uncertainty that financial analysts, for their part, do not mask. Wang Feili, an industrial analyst at UBS Securities China, estimates that 'brain technology' remains the main bottleneck for the commercialisation of humanoids, and that an 'electric car moment' for the sector is improbable in the next five years. As long as actual use remains thin, the only available competition ground is the figure — and the figure, when it is not arbitrated by anyone, is manufactured.
The calculator rather than the crown
After the demo, after the word, after the scene, it is therefore the number that lies. Three 'number one in the world' claims mutually exclude each other in the same week; the man who cannot produce any production figure promises the biggest figure of all. The only thing truly verifiable in these six days is not a ranking: it is an admission — that of Musk acknowledging that none of his robots is working. The rest is a matter of competing spreadsheets.
The formula this thread carried since December — do not ask if the robot is autonomous, ask who holds the stopwatch — goes up a notch. This week, the stopwatch became a calculator, and each brought its own.