---
title: "Galbot (银河通用) closes over $300 million raise and a 1,000-robot factory order — humanoids move from demo to deployment"
date: 2026-08-11
category: Capital & Markets
site: NeuroAI
canonical: https://neuroai.site/a/na-galbot-300m-robot-factory
language: en
---

# Galbot (银河通用) closes over $300 million raise and a 1,000-robot factory order — humanoids move from demo to deployment

> Galbot (银河通用), a Beijing humanoid-robot developer, raised more than $300 million (about 2.16 billion yuan) at a valuation above 20 billion yuan (about $2.78 billion / HK$21.7 billion), and signed a deal to deploy over 1,000 humanoid robots at precision manufacturer Parker Precision (派克精密). The twin announcements signal factories are shifting from curiosity to purchase orders.

The first week of August lit up China's robotics circles with two headlines at once. Galbot (银河通用), a Beijing-based developer of humanoid robots, closed a new financing round of more than $300 million (about 2.16 billion yuan) — a sum the company says sets a record for a single round among humanoid-robot firms — pushing its valuation past 20 billion yuan (about $2.78 billion / HK$21.7 billion). Almost simultaneously, Galbot struck a strategic partnership with Parker Precision (派克精密), a precision manufacturer, to deploy more than 1,000 humanoid intelligent robots in industrial precision manufacturing. The combination of a giant raise and a four-figure order marks a shift: downstream factories are moving from "let's take a look" to "let's place an order."

For readers new to the space: a "humanoid robot" is a machine built roughly in human form, and the Chinese industry has spent two years producespectacular demo videos — backflips, folding clothes, pouring coffee. The harder question is whether any of them can actually hold a job on a factory floor. Galbot's announcements are a data point that the answer is starting to be yes.

## Key takeaways

- Galbot raised **more than $300 million** (about 2.16 billion yuan), with a valuation **above 20 billion yuan** (about $2.78 billion / HK$21.7 billion).

- The Parker Precision (派克精密) deal calls for **over 1,000** humanoid robots in precision manufacturing.

- Galbot's S1 model has run **7×24** autonomous shifts for **over 3 months** on a CATL (宁德时代) production line.

- Confirmed deep-pay customers include **CATL, Bosch (博世), Toyota (丰田), and BAIC (北汽)**.

- China's Ministry of Industry and Information Technology (工信部) projects **over 100,000** humanoid units produced in 2026, from 230+ firms shipping 400+ models.

## What a thousand-unit order really means

Building one trade-show prototype and delivering a thousand units onto a production line are two entirely different species. Parker Precision supplies core components for automobiles and home appliances, where precision tolerates near-zero error, and where multi-variety, small-batch lines demand general intelligence from the robot. A thousand-unit deployment means the technical package — cost, reliability, maintainability, and ability to deliver in volume — has passed the test of real industry.

Galbot's approach is pragmatic rather than flashy. Instead of obsessing over mimicking human legs, it uses a "foldable torso plus wheeled base" configuration, trading bipedal mimicry for the mobility efficiency and stability factories actually care about. On the software side, it builds a training set in the hundreds of billions of tokens primarily from simulation, then attacks fine manipulation with a dexterous-hand neurodynamic model. This is engineering, not theatre.

## From showboating to line delivery

In the first half of 2026, the narrative center of gravity in embodied AI shifted from the "hundred-model war" — a reference to the explosion of Chinese large-language-model startups — to "mass production and delivery." China's Ministry of Industry and Information Technology (工信部, the MIIT) has stated that 2026 humanoid-robot output could exceed 100,000 units, with over 230 companies having released more than 400 complete machines. As Unitree (宇树科技) moved toward a Star Market (科创板, Shanghai's tech-board) public offering and Galbot landed its thousand-unit order, the industry's main thread flipped from "who can backflip" to "who can work steadily."

For founders, the signal is clear: the second half of the robotics race treats hardware as merely the entry ticket. What真正 separates winners is the "brain" and deep scenario expertise. The suggestion for smaller teams is to chase not the headline of building a whole machine, but the unglamorous integration, operations, and data-feedback work behind line-level deployment — that is where the real cracks to slip into exist. The sexiest part of embodied intelligence was never the demo video; it is the production line still running at 2 a.m.

## Why capital is willing to bet

A single round above $300 million and a valuation past 20 billion yuan reflect a simple logic: capital backs whoever first closes the commercial loop of "scale and autonomous work." Galbot's confidence rests not only on the raise but on real customers — CATL (宁德时代), Bosch (博世), Toyota (丰田), and BAIC (北汽) are already in deep cooperation, and its Galbot S1 has run 7×24 autonomous work for over three months on a CATL line. Orders beat slides; that is the hardest proof of line delivery.

The path offers a second lesson for teams entering the field: don't start by trying to build a whole machine; first become a reliable component or scenario solution that gets integrated. While giants fight over "who IPOs first," a smaller player can be the screw no thousand-unit deployment can do without — and that is steadier. The golden decade of embodied intelligence belongs both to the front-stage machine stars and to the back-stage suppliers who push reliability to the extreme.

## Honest limitations

This article draws on a Chinese tech-outlet report and on figures the outlet attributes to Galbot and its partners; the financing amount, valuation, and order size are company-stated and were not independently audited at the time of writing. The claim of a "single-round record" among humanoid firms is the company's characterization and should be treated as a marketing claim pending verification against other raises. The "over 3 months of 7×24 autonomous operation" on a CATL line is reported as a deployment fact but lacks a public, third-party uptime audit. MIIT's 100,000-unit 2026 production projection is a government estimate, not a guaranteed outcome, and the 230-firm / 400-model counts are industry tallies that shift constantly. Dollar and Hong Kong-dollar conversions use approximate rates (1 USD ≈ 7.2 RMB; 1 HKD ≈ 0.92 RMB) and are indicative only. No investment advice is intended.

## Sources

Based on reporting by the Chinese tech outlet Naojiwang (脑机派对), 2026-08-11, citing Galbot (银河通用) disclosures, the Parker Precision (派克精密) partnership announcement, and statements attributed to China's Ministry of Industry and Information Technology (工信部).

---

Published by NeuroAI (https://neuroai.site/) — https://neuroai.site/a/na-galbot-300m-robot-factory
Free to quote with attribution and a link to the original.
