Physical AI Pulled In $47.4B in H1 2026: Four Times the Money, a Flat Deal Count

Crunchbase data shows that in the first half of 2026, physical AI companies — robotics, drones, aerospace and autonomous machines — raised $47.4 billion across 521 deals. In the second half of 2025 the same cohort raised $12 billion across 470 deals. Capital rose almost fourfold; the deal count rose by about a tenth.

Four times the money, but not a wider door

Dividing those same two figures, the average round size in the sector moved from roughly $26 million to roughly $91 million in six months. In other words, the new capital is not creating room for new companies; it is going to names already on the list. For angel investors that distinction matters: the heat is at later stages, not at seed.

Within robotics alone, Crunchbase counted $18.8 billion invested in 2026 as of its mid-June report, against $15 billion for all of 2025 and $14.1 billion in the 2021 funding peak. A sector long viewed as expensive and slow to return capital is being repriced.

Valuations running ahead of revenue

In March 2026, 37 companies joined the Crunchbase unicorn board — the highest monthly count in close to four years — with robotics leading at six names, three of them Chinese. The largest round that month was the $1 billion raise by Advanced Machine Intelligence, Yann LeCun's Paris-based lab. In June, another 34 companies joined, adding more than $110 billion in value; ten frontier labs accounted for $65 billion of that, with DeepSeek the highest at $50 billion.

Hardware does not run on software's clock

• Capital intensity: money goes into tooling, components and pilot lines — spending that is not recoverable if the product pivots.
• Supply chains set the schedule: component lead times, not coding speed, determine launch dates.
• Gross margins are lower than in software, so the same revenue supports a smaller valuation at the next round.
• Acquirers are mostly industrial groups, so due diligence runs longer and leans on certification, safety and engineering documentation.

The context matters: per Crunchbase, global startup funding hit a record $510 billion in the first half of 2026, and $42 billion in August alone, up 122% year over year. Physical AI is riding a cycle larger than itself. When that cycle slows, the most cash-hungry cohort feels it first.