The humanoid robotics sector raised $4.2 billion across 34 disclosed funding rounds in 2025, according to HumanoidIntel's full-year funding tracker — more than double the $2.1 billion raised in 2024 and roughly 14x the $300 million raised in 2022. The acceleration reflects both genuine technology progress and the classic late-stage venture pattern of institutional capital concentrating in markets perceived to be approaching inflection.
By Geography: US vs. China, Round by Round
US-based companies captured 52% of 2025 humanoid funding ($2.18B), with Chinese companies taking 38% ($1.60B) and European companies the remaining 10% ($420M). The US lead is primarily a function of two mega-rounds: Figure AI's $1.5B Series C and Apptronik's $350M Series A, both of which closed in late 2025 or very early 2026.
At the round-count level, the picture inverts: Chinese companies completed 19 of the 34 disclosed rounds, compared to 10 for US companies and 5 for European. Chinese investment is more distributed across the humanoid ecosystem — seeding a dozen companies at Series A and B stage — while US investment is more concentrated in a handful of leaders.
2025 Funding by Geography
| Region | Capital | Rounds | Largest Round | |--------|---------|--------|---------------| | USA | $2.18B | 10 | Figure AI $1.5B (Series C) | | China | $1.60B | 19 | Agibot $200M (Series B) | | Europe | $420M | 5 | NEURA Robotics $120M (Series C) |
By Stage: The Series A Bottleneck
The most important structural feature of 2025 funding is the concentration at Series A ($50M-$150M). Thirteen of the 34 rounds fell into this category, reflecting a cohort of companies that prototyped successfully in 2022-2023 and spent 2024-2025 converting prototypes into early pilots.
Series B rounds — indicating companies with pilot deployment data moving toward production — were notably sparse: only 6 disclosed rounds at this stage in 2025. This is the stage where unit economics must hold, safety cases must be made, and enterprise sales cycles must be navigated. The bottleneck at Series B suggests that most of the 2022-2023 cohort is still converting pilot evidence into scalable commercial contracts.
This matters for 2026 projections. The companies that successfully navigate Series B in 2026 will be the commercial leaders of 2027-2028. The funding pipeline implies 8-12 potential Series B candidates, but historical conversion rates suggest 3-4 will succeed on compelling terms.
The Emergence of AI Foundation Model Companies
A qualitatively new category emerged in 2025: pure-play robotics AI companies that raise at valuations comparable to hardware companies without manufacturing a single robot. Physical Intelligence ($470M+ at $2.4B valuation) and Skild AI ($300M at $1.5B valuation) together raised $770M in 2025 at a combined implied valuation of $3.9B.
These companies are betting that the bottleneck in humanoid deployment is not hardware — which is commoditizing, as Unitree's $16,000 H1 demonstrates — but the intelligence layer. If that thesis is correct, the most valuable humanoid companies in 2030 may not manufacture robots at all, instead licensing foundation models to hardware OEMs much as Qualcomm licenses modem IP to smartphone manufacturers.
The risk to this thesis is vertical integration: if Tesla, Figure AI, and Apptronik all develop proprietary AI stacks (as each is attempting), the market for third-party robot AI models may be smaller and more price-competitive than the current valuations assume.
What 2026 Looks Like
HumanoidIntel's forward estimate for 2026 humanoid funding is $5.5B-$7.0B, with the following catalysts:
- Tesla external sales announcement (expected H1 2026): Will trigger a rerating of sector valuations and likely catalyze several new entrants raising at elevated prices.
- Chinese state capital acceleration: The PRC Ministry of Industry and Information Technology's 2025 humanoid roadmap targeted 10 domestic "champion" companies by 2027. State-guided funds are expected to deploy $800M-$1.2B into the sector in 2026.
- First Series C rounds for Chinese leaders: Agibot, Kepler, and Astribot are expected to pursue Series C financings in 2026 as they convert pilot programs into production contracts.
- European institutional catch-up: NEURA's Series C featuring all three German automotive OEMs as investors opened the door for European corporate venture to participate more systematically. An estimated $400-600M in European humanoid rounds is expected in 2026.
The macro risk is interest rate and risk appetite normalization. The 2025 surge was partially enabled by declining interest rates and recovering tech valuations broadly. A reversal of either trend would compress multiples and slow the pace of Series B and C activity.
FAQ
Which investor was most active in humanoid robotics in 2025? Sequoia Capital (both US and China funds) participated in the most rounds by count — seven confirmed investments across hardware companies and AI foundation model companies. By capital deployed, SoftBank Vision Fund 2 led with over $400M across Skild AI and strategic positions in Unitree and UBTECH.
Are any humanoid companies profitable? As of end-2025, no pure-play humanoid hardware company has disclosed profitability. Fourier Intelligence, which derives substantial revenue from rehabilitation exoskeleton sales, is the closest to break-even. The general expectation among investors is that unit economics turn positive at approximately 500-1,000 units of sustained annual production volume — a threshold that only Tesla has crossed, internally.
What percentage of 2025 funding was in Chinese companies? Chinese companies captured approximately 38% of total disclosed humanoid funding in 2025 ($1.6B of $4.2B). When including undisclosed state capital estimated by HumanoidIntel, the Chinese share rises to approximately 45-48%.