Robotics Startup Funding Reaches $9.9 Billion in Q3 as US and China Lead

Robotics startups attracted $9.9 billion in global venture funding in Q3 2026, more than twice the $4.3 billion recorded in Q3 2025, according to Crunchbase data1. The concentration of capital in the United States and China points to where investors are backing companies that build robots, models and the infrastructure needed to make machines work in physical settings.

Funding across the sector from Q1 through Q3 2026 is up around 185% year over year. The latest quarter extends a run in which robotics companies have raised nearly $10 billion or more per quarter. Those totals show the scale of current financing, but they don't establish how quickly the companies can turn capital into products or lasting businesses.

China and the US account for most of the funding

China received $13.2 billion in robotics investment this year, while U.S.-based companies received $12.1 billion. Germany ranked third, at $1.7 billion—a wide gap that underscores how concentrated funding is among the top two markets.

Two of China's largest rounds went to companies developing robots and the systems behind them. Guangzhou-based Xpeng Robotics, which builds AI models and physical robots, raised $900 million. Shenzhen-based AI² Robotics, whose work spans manufacturing, healthcare and services, raised a $736 million Series C round.

In the U.S., Los Angeles-based Travis Kalanick's Atoms raised a billion-dollar round. Its businesses include a ghost kitchen operation alongside robotics divisions focused on mining and transport, making it a broader mix than a single-purpose robotics company. Pittsburgh-based Skild AI raised $1.4 billion for Skild Brain, its robotics foundation model.

Germany's Neura Robotics also raised $1.4 billion this year. The company is developing cognitive robots supported by AI and data infrastructure. Its round stands out against Germany's $1.7 billion total for the sector, although the source doesn't specify how that national total is distributed across companies or rounds.

Physical AI requires data from the physical world

The funding surge is tied in part to a shift from AI systems trained on online text toward models that must interpret surroundings and act through machines. “We’re just seeing physical AI take off,” said Ulrik Stig Hansen, co-CEO of San Francisco-based Encord, a provider of data tools for physical AI. Hansen spoke to Crunchbase News at the company's inaugural Long Horizon Physical AI Summit in San Francisco earlier this month.

For text-based models, large volumes of training material are already available online. Physical AI developers have to collect information from the environments where robots operate, then translate sensor inputs into actions. “Physical AI is very different because there is no internet of physical AI data, and so the data first has to be collected, and you have to convert it from sensor tokens to action tokens,” Hansen said.

That work can involve vision, touch, force and torque data. Collecting and storing those different inputs, then curating, enriching and validating them, adds cost and time before a model can be trained and used reliably. “Storing all that data, curating all that data, enriching all that data, and validating that all the data works for these systems is a very expensive operation,” Hansen said.

Hansen described the field as still being in a pre-training phase, with a post-training phase approaching. That points to a long development path: large funding rounds can support data collection, computing and model development, but the source doesn't establish when those investments will produce commercial returns.

More companies are attracting capital

Around 75 robotics companies on the unicorn board have raised funding in the past three years. Forty-five—more than half of that group—joined the board this year. The newly valued companies span data and model development as well as applications in heavy industry, manufacturing, logistics and household settings.

That breadth helps explain why Hansen sees room for multiple companies to succeed. “There are a lot of companies that are raising capital, and I think a lot of them are going to win, just because there’s so many different types of applications to the physical world,” he said.

The range of uses also makes the funding figures difficult to read as a single contest. Companies building general-purpose models face different development needs from those applying robotics to a specific workplace or task. The figures show substantial investor interest across the sector, while the source leaves open how many approaches will find repeatable demand and support their costs.

For founders and workers, the practical question is whether this capital can move projects from data and model development into dependable use. For investors and customers, the mix of companies and applications means that progress in one segment may not predict progress in another. The next phase will test whether the sector can turn its funding surge into systems that work across the settings they are designed for.

Related Crunchbase queries

• Global Physical AI Venture Funding In 2026
• Global Robotics Venture Funding In 2026

Related reading

• Sector Snapshot: Robotics Startups On Fire As Venture Funding Surges To Record Numbers In 2026
• Embodied AI Fuels Record Robotics Funding In China As IPO Momentum Builds
• VCs Pour Billions Into Physical AI As The Next Wave Of AI Investing Takes Shape

Illustration: Dom Guzman