On 9 December 2025, AgiBot celebrated in grand style the rollout of its 5000th humanoid robot. The scene, broadcast via livestream from the Chinese manufacturer’s Shanghai factory, had all the hallmarks of an industrial milestone: mass production, finally achieved. But the detail that should have gone unnoticed tells a different story. This 5000th unit did not head to a factory, warehouse, or store. It was handed over, symbolically, to the studio of Chinese actor Huang Xiaoming.

A showcase robot meant to embody the transition to scale, ending up in show business: the image is too good to ignore. And when you look closely at the composition of the figure itself, the narrative of the ‘5000th worker’ cracks.

The 5000th robot is a stage model

According to Pandaily, unit number 5000 was a Lingxi X2, delivered symbolically to Huang Xiaoming’s studio. The choice is anything but trivial. The X2 belongs to the lightest range in AgiBot’s catalogue: an agile, half-size biped, designed for demonstration and entertainment far more than for handling.

In other words, to celebrate a milestone in industrial manufacturing, the company chose its most spectacular robot and entrusted it to a figure of the performing arts. The casting of the event is, in itself, a confession: they are putting the robot forward where it performs — the stage — and taking it away from where it should, one day, work.

What AgiBot actually produces

The detail of the produced fleet, reported by Interesting Engineering, confirms this reading. Of the 5000 units, the breakdown by range is eloquent:

  • X-Series (agile half-size biped, entertainment-oriented): 1846 units — the most produced range.
  • A-Series (full-size humanoid): 1742 units.
  • G-Series (wheeled base, task-oriented): 1412 units.

The observation is clear: it is neither the full-size industrial humanoid nor the task-oriented model that dominates production, but the entertainment range. The most manufactured chassis is not the one supposed to replace an operator on a line or an employee in a store aisle. It is the one that knows how to dance.

Manufacturing 5000 robots is not deploying 5000 workers

This is where the entire slippage lies. Producing 5000 humanoids is a genuine manufacturing feat — assembling 5000 chassis, feeding a line, reaching a volume. But this manufacturing success is presented as a work success: 5000 robots about to enter service. Yet nothing in the staging supports this equivalence. The figure that matters is not ‘5000’. It is which of the 5000, and where they end up.

AgiBot’s commercial strategy reinforces this doubt. According to the Japanese media outlet Robotstart, the company prioritises the development of a partner ecosystem over the direct sale of robots: it provides open-source development platforms, data, and large language models, and seeks local partners to train its machines for cultural and commercial uses. A company that mastered ready-to-work robots would sell them turnkey. AgiBot, however, sells tools to learn how to make them work.

The real work is for later

The company’s discourse actually acknowledges this deferral. Again according to Robotstart, AgiBot’s marketing director, Qiu Heng, describes a three-phase deployment strategy, classified by complexity:

  • Simple environment, simple task: factories and warehouses.
  • Complex environment, simple task: supermarkets.
  • Complex environment, complex task: domestic work, presented as a potential future market of hundreds of millions of units.

This final phase, the most ambitious, remains explicitly in the future. Qiu Heng also cites Japanese konbini — these convenience stores open 24 hours a day — as a specific opportunity, due to high labour costs and security risks weighing on night shifts. An opportunity, therefore; not yet a deployment. The executive presented the A2, X2, and G2 ranges at the company’s stand at the iREX2025 trade show in Tokyo, but the transition from projection to contract remains to be written.

A deficit disguised as success

The 5000-robot milestone is real, and it says something true: the sector has learned to produce humanoids quickly. Very quickly. But it has not yet learned to make them do real work at the same pace. By choosing a stage robot as a symbol and entrusting it to an actor, AgiBot inadvertently illustrates this gap. The first success — manufacturing — serves to mask the second deficit — employment. It remains to be seen, when the lines are running at full capacity, how many of these robots will actually go to the factory, and how many will continue to go to the studio.