Figure AI reportedly building humanoids at 1/hour
- Figure said on April 29 it has pushed Figure 03 output at its BotQ factory from one robot per day to one per hour. - The company says it has built more than 350 robots, hit a 24x throughput jump in under 120 days, and targets 55 units this week. - That matters because humanoids are shifting from demo-stage robots toward factory-style manufacturing — while Figure is already valued at $39 billion.
Humanoid robots are finally crossing a line people in tech have been promising for years. Not intelligence alone — production. Figure AI said this week that its BotQ factory in California has gone from building one Figure 03 robot per day to one per hour in under 120 days, with more than 350 units built so far. That does not mean humanoids are suddenly everywhere. But it does mean one of the most closely watched startups in the category is trying to prove this is now a manufacturing problem, not just a lab problem. (figure.ai) ### What actually changed? The concrete news is simple. On April 29, Figure posted a production update saying its hardware and manufacturing teams had transformed BotQ from a plan into a working high-output line. The company said throughput improved 24x — from one Figure 03 per day to one per hour — and said the factory is on pace to build 55 robots this week. (figure.ai) ### Why is one per hour a big deal? Because the bottleneck in humanoids has never just been “can the robot move?” It has been “can you build enough of them, with consistent quality, at a cost that makes deployment real?” A robot that looks great in a demo is not a business. A robot that can come off a line every hour starts to look like the firs(figure.ai)y yield is 99.3%, and it has produced more than 9,000 actuators — all signs the company is working the boring factory problems that usually decide whether a hardware category lives or dies. (figure.ai) ### Is this already mass production? Not really — at least not in the way cars or phones are mass produced. One robot per hour is still early industrial scale. If that pace held nonstop for a year, it would imply something like 8,700 units annually, and Figure’s own March 2025 BotQ announcement framed the first-generation line as capable of up t(figure.ai)ve arrived” and more like “the ramp is starting to match the original factory thesis.” (figure.ai) ### Why are people tying this to valuation? Because investors have already priced Figure like a company that could matter at enormous scale. In September 2025, Figure said it raised more than $1 billion in Series C funding at a $39 billion post-money valuation. That is a huge number for a company still proving deployment volume. So every manufacturing milestone now get(figure.ai)hey build robots?” but “can they grow into that valuation?” (figure.ai) ### Does this validate the trillion-dollar talk? Not directly. The social posts making the rounds are bundling together three different stories — humanoid production ramps, agent software commercialization, and frontier-model valuation pressure. Those things are related, but they are not the same business. Figure’s update supports the first claim: humanoid output i(figure.ai)elf prove that robotics companies, or AI labs tied to them, deserve trillion-dollar expectations. That leap is still an investor bet. (figure.ai) ### What is the real catch? Deployment. Building a robot is one hurdle. Getting customers to use fleets of them productively is the harder one. Figure has talked publicly about BotQ, BMW work, and broader commercial ambitions, but the market still needs evidence that these machines can operate reliably enough, cheaply enough, and long enough to (figure.ai)the verdict. (figure.ai) ### So what should you take from this? The important part is not the exact one-hour cadence. It is that a leading humanoid startup is now talking like a manufacturer and publishing manufacturing metrics instead of just robot videos. That is a real shift. But it is still the beginning of the scale story, not the end. (figure.ai)
