Tokenta Signal SIGN IN TERMINAL >
2026.11.07 13:37 UTC
● ISSUE 256 · MICROSOFT × ORACLE × AI CLOUD · Reuters report of Bloomberg News (Duke FM syndication)● ISSUE 255 · TERAWULF × Q3 2026 × WULF · TeraWulf Q3 2026 results (GlobeNewswire release)● ISSUE 254 · PSI × QUANTUM · PSI Aaronson-Ambainis proof announcement (PSI)● ISSUE 253 · IBM × QUANTUM · IBM Quantum System Two Chicago unveiling (PR Newswire)● ISSUE 252 · OKLO × Q3 2026 · Oklo Q3 2026 Form 10-Q (SEC)● ISSUE 251 · VISTRA × Q3 2026 × VST · Vistra Q3 2026 results (PR Newswire)
● ISSUE 256 · MICROSOFT × ORACLE × AI CLOUD · Reuters report of Bloomberg News (Duke FM syndication)● ISSUE 255 · TERAWULF × Q3 2026 × WULF · TeraWulf Q3 2026 results (GlobeNewswire release)● ISSUE 254 · PSI × QUANTUM · PSI Aaronson-Ambainis proof announcement (PSI)● ISSUE 253 · IBM × QUANTUM · IBM Quantum System Two Chicago unveiling (PR Newswire)● ISSUE 252 · OKLO × Q3 2026 · Oklo Q3 2026 Form 10-Q (SEC)● ISSUE 251 · VISTRA × Q3 2026 × VST · Vistra Q3 2026 results (PR Newswire)
ROBOTICS & AI · SIGNAL REPORT

$13 Billion in Half a Year: China's Robots Fix Their Brains Before Their Moves

Reuters reported on November 2 that Chinese embodied AI companies raised RMB 93.5 billion, about $13 billion, in the first half of 2026 across 322 deals, five times a year earlier, with funding flowing to robot intelligence developers rather than only complete-machine makers.

ISSUE 147 · 2026-11-01 · SOURCE / Reuters report, republished · STATUS / VERIFIED · TOKENTA SIGNAL RESEARCH
CHINA × EMBODIED AIREPORTED + ANALYSIS
IMAGE: Reuters report, republished — FROM THE ORIGINAL REPORT ↗
RMB 93.5B
H1 2026 FUNDING
The amount raised by Chinese embodied AI companies in the first half of 2026.
322
DEALS
The number of deals in the first half, five times the same period a year earlier.
~$13B
IN DOLLARS
The funding amount in approximate dollar terms.
KEY TAKEAWAYS

WHAT HAPPENEDWhat Happened

Reuters reported from Beijing on November 2 that China's robotics companies are shifting their focus from movement to brains. In recent years these companies built humanoid robots that can run, dance and do backflips; now they must solve the software that lets robots generalize to unfamiliar tasks.

The money has already moved first. Data from ITjuzi show Chinese embodied AI companies raised RMB 93.5 billion, about $13 billion, in the first half of 2026, across 322 deals, five times the same period a year earlier. Funding is flowing increasingly to specialist developers of robot intelligence, not only to complete-machine makers.

Reuters interviewed founders and executives at five Chinese robotics companies. The consensus is that hardware demonstrations have proven the machines can move; the real threshold is understanding environments, learning unfamiliar tasks and adapting to new work without being programmed item by item.

MOVES DONE, BRAINS NEXTAfter Movement Comes Work

Running, jumping and somersaults are demonstrations; work is a product. A demonstration only needs the movements choreographed; work requires judgment in environments never seen before. One is a choreography problem, the other a generalization problem, and the difficulty is not of the same order.

Chinese makers built the movement first and are adding the brains after, for practical reasons. The hardware supply chain and manufacturing capacity already exist, while the intelligence layer needs time and data. Money now flooding into the intelligence layer shows the industry itself knows where its short board is.

MONEY FOLLOWS SOFTWAREFivefold Growth, and the Money Is Buying Software, Not Machines

The most instructive part of the RMB 93.5 billion figure is not the total but the structure. Funding is flowing increasingly to specialist developers of robot intelligence, which shows investors are no longer satisfied with backing yet another company whose robots can walk.

Complete-machine companies and intelligence companies are dividing the labor. The former keep driving hardware costs down; the latter focus on making the same machine adapt to more tasks. The finer the division of labor, the more the ecosystem resembles today's phone industry, with its own leaders in hardware and software alike.

THE GENERALIZATION TESTGeneral Tasks Are Embodied AI's Real Examination

The promise of embodied AI has always been that one machine can learn tasks it was never programmed for item by item. Until that promise is delivered, every funding figure is only an entry ticket.

Reuters' report gives no timetable for delivery, and the judgment of the five interviewed companies is only a directional consensus. The next round of embodied AI news should not be another funding total, but a machine that worked in a real setting for a month straight without incident.

THE SHIFTWhat Changed

FROM DEMOS TO JOBS
From movement demonstrations to task generalization
Choreographed performances give way to work never seen before.
FROM HARDWARE TO BRAINS
From backing complete machines to backing the intelligence layer
After the hardware dividend, software became the scarce good.

SIGNAL JUDGMENT — ANALYSISRobots Fix Their Brains Before Their Moves

Tokenta Signal's judgment: the embodied AI race is changing tracks. The dividend from movement demonstrations is spent; the next stage is about whose machines can learn to work on unfamiliar tasks by themselves.

WHAT REMAINS UNCERTAINWhat Remains Uncertain

The limits of this report: the RMB 93.5 billion and 322 deals are figures from ITjuzi provided to Reuters, with the industry boundary defined by that firm; different statistical scopes produce different numbers. Reuters' assessment is based on interviews with founders and executives at five Chinese robotics companies, not an industry-wide census.

EVIDENCEEvidence

The original confirms Reuters' November 2 report that Chinese embodied AI companies raised RMB 93.5 billion, about $13 billion, in the first half of 2026 across 322 deals, five times a year earlier, with funding flowing to robot intelligence developers.

ORIGINAL: China's robot makers have mastered the moves. Now comes the race for brains ↗
wncy.com · 2026-11-01 · Reuters report, republished

PRIMARY SOURCESPrimary Sources

RELATEDRelated Intelligence

CORRECTIONS: if a fact on this page is wrong, we keep a correction record on the page. Contact team@tokenta.space