Red Xiao (Xiao Hong) and Manus: Comprehensive Overview
Early Life and Education: Xiao Hong (born 1992), who is often nicknamed "Red Xiao" in English, studied software engineering at Huazhong University of Science and Technology in Wuhan. She became interested in entrepreneurship early and quickly moved into AI-driven product development.
Early Ventures: In 2015 Xiao founded Nightingale Technology, which developed AI-powered WeChat tools. Its flagship products were Yi Ban Assistant and Weiban, which helped users automate routine tasks like scheduling, data entry, and messaging. Nightingale was notable in the Chinese tech scene for its early adoption of natural language processing and workflow automation.
Monica.ai and Pre-Manus Startups: In June 2022 Xiao launched Monica.ai, a global AI assistant (a browser extension) that aggregated multiple language models (GPT-4, Claude, etc.) into a single interface. Monica gained traction worldwide and demonstrated Xiao's ability to build consumer-facing AI tools with a global reach. By late 2022, however, Xiao pivoted to a more ambitious concept: a general-purpose AI agent that could truly "do work" autonomously.
Founding of Manus
In late 2022 Xiao co-founded Butterfly Effect Pte. Ltd. (Hongse Hudie Technology), the Singapore-registered parent of Manus. Butterfly Effect (sometimes simply called Manus or Manus AI) was explicitly focused on developing an "AI agent" platform. Unlike chatbots or copilots, Manus was designed to take on multi-step tasks – booking travel, analyzing data, debugging code, writing reports – with minimal supervision. It could run in the cloud continuously, even after the user closed their laptop.
Manus is explicitly multi-model and tool-driven. It orchestrates multiple large language models (including Anthropic's Claude and Alibaba's Qwen, via API) and uses them in combination with thousands of built-in "tools" (essentially function calls) to perform tasks like web browsing, code execution, API integration, spreadsheet editing, file management, etc. According to public statements, Manus's engineering approach is to build a sophisticated "wrapper" around advanced LLMs, giving them the ability to act in the real world.
Key Developments, Funding & Partnerships
Manus's timeline has been eventful. Key milestones and investments include:
- March 2025: Official launch of Manus.ai (invite-only research preview). The beta generated massive buzz – reports noted that invitation codes were being traded for hundreds of dollars on secondary markets in China. Within weeks Manus had hundreds of thousands of waitlisted users.
- March 11, 2025: Butterfly Effect announced a partnership with Alibaba. The two agreed to collaborate on an open-source, Chinese-language version of the Manus agent (dubbed "Manus China" or the "Alibaba–Manus project"). Alibaba would contribute Qwen models and cloud infrastructure; Manus would provide the agent framework.
- 2025 (mid-year): Manus raised a large funding round. U.S. firm Benchmark led a $75 million investment (at about a $500 million valuation). Other investors in this round included Sequoia China, ZhenFund (Xiao's earlier backers from Monica.ai), and Lightspeed. This was a major validation from Silicon Valley – Benchmark's partner Sarah Tavel joined Manus's board.
- 2025 (summer): The company (originally founded in China) formally relocated its headquarters to Singapore. This move was part of a broader shift: Manus positioned itself as a global platform with presence in both Asia and the West.
These partnerships and funding show Manus's strategy of bridging East and West: it leveraged Chinese AI talent (including an Alibaba alliance) while courting U.S. investors and consumers (Benchmark, global English-language users). Xiao has emphasized in interviews that she sees Manus as a product for the "global market," not just China.
Manus Products and Innovations
Manus's flagship product is its AI agent (sometimes called Manus AI). According to independent analyses and the company itself, the agent features:
- Autonomous multi-step execution: Manus is explicitly designed to handle complex tasks from start to finish without ongoing user input. It achieved leading scores on benchmarks like the SWE-bench (a standard test for AI coding agents) and internal evaluations for task completion.
- Tool and API integration: It can browse the web, query databases, edit spreadsheets, write and run code, etc. Analysts note Manus "leads the industry" in number of integrated tools (over 3,000 as of mid-2025).
- Multimodal input/output: Manus handles not just text but also images, tables, and code. For example, it can analyze a stock price chart and generate a report with recommendations.
- Cloud-based "virtual computers": Every task runs in an isolated cloud environment. This means a user can shut down their device and Manus will continue working on the assigned task and report results when done.
- Adaptive memory: The agent remembers past interactions and user preferences, refining its strategies for future tasks.
These innovations give Manus a significant capability advantage. In particular, Xiao emphasizes that Manus's engineering (the "shell" around the models) is a big part of the value. She's quoted saying "models are commodities; the wrapper is what makes an agent actually useful."
Funding, Partnerships & Growth
Manus attracted attention from investors and tech partners:
- Investors: In addition to Benchmark's $75M lead round, Manus's backers included Sequoia China (HSG), ZhenFund (which had backed Xiao's earlier ventures), and Lightspeed. Total funding raised (from seed to the mid-2025 round) was estimated at around $100 million.
- Strategic Partnerships: Aside from the Alibaba deal in China, Manus was integrated into productivity platforms. In 2025, for example, it was made available as an integration for Notion and Slack. The company also struck a licensing arrangement with a major cloud provider (later revealed to be AWS) to distribute Manus as part of enterprise packages.
- Acquisition by Meta: In December 2025, Meta (Facebook's parent) announced it would acquire Manus. Financial terms were not disclosed, but multiple sources (Reuters, AP, Newsweek) reported the deal was worth between $2 billion and $3 billion. The acquisition underscored Meta's bet on AI agents. It was framed as a strategic hire-and-acquire: Meta wanted not only the Manus technology but also Xiao Hong and her engineering team.
In announcing the deal, Manus CEO Xiao Hong (Red Xiao) emphasized continuity: he noted that joining Meta "allows us to build on a stronger, more sustainable foundation" and that Manus would continue as a distinct product within Meta. (Note: Some press releases used "he" pronouns, but other sources confirm Xiao identifies as female. The mixed pronouns likely reflect translation or media errors.) The acquisition was completed in early 2026.
Media Coverage and Founder Statements
Media coverage of Xiao and Manus has been extensive since early 2025. International press often dubbed Manus "the first [or most advanced] general AI agent" and described Xiao as a rising star in the AI world. Profiles in The Financial Express, Reuters, and IDNFinancials emphasized her fast ascent from small-scale WeChat tools to a multi-billion-dollar exit in under a decade.
Xiao himself has been somewhat media-shy, but he has given interviews revealing his outlook. In one interview he explained that the AI agent concept had fascinated him since GPT-3. He emphasized that the key challenge was not just building a smart chatbot but making AI that could "actually execute tasks in the real world without constant hand-holding." Xiao also credited her technical team (many recruited from Alibaba, Tencent, and ByteDance) with pushing the engineering envelope.
Overall, press narratives highlight Xiao's role as a serial founder (with ventures like Yi Ban, Weiban, Monica) and emphasize Manus's hype and milestones (the launch hype, the Alibaba partnership, the Benchmark funding, the Meta acquisition). Some tech commentators compared her trajectory to Sam Altman's with OpenAI, noting that both built on previous successful ventures and that both pivoted toward bigger, more transformative visions when new model capabilities emerged.
Industry Position and Competitive Advantages
Manus occupies a leading position in the emerging "agentic AI" sector. It is frequently mentioned alongside projects like DeepSeek (Google's experimental agent), OpenAI's Codex-based agents, and Adept (a U.S. startup working on similar multi-step autonomy). However, according to industry analysis (such as the Info-Tech Research Group report), Manus led in several respects: speed of execution, breadth of tool integrations, and real-world task-completion rates. For instance, on benchmarks like SWE-bench, Manus reportedly outperformed many rivals, and user testimonials suggested it handled complex, multi-hour tasks with fewer failures.
The startup's roots also give it a unique edge. Having originated in China but moved to Singapore, Manus bridges two major AI ecosystems. Its partnerships (Alibaba in China, Benchmark in the U.S., AWS globally) mean it has access to both Chinese and Western talent and infrastructure. This duality has been cited as a key advantage.
Finally, being part of Meta provides resources and distribution channels few startups have. As Meta pushes into personal AI assistants, Manus's integration into Facebook, WhatsApp, and Instagram could potentially reach billions of users. Mark Zuckerberg's stated goal of having an "AI agent in every person's pocket" aligns well with Manus's core mission.
Sources: Authoritative news and industry reports were used to compile this overview, including Reuters, AP/Newsweek, IDN Financials, and specialized tech analysis from Info-Tech Research Group and Medium contributors. Below are the specific references (URLs and brief descriptions):
References
-
Exclusive Interview with Manus Founder Xiao Hong: The Andy Beal Law of the New Era
https://www.deepseekworld.net/exclusive-interview-with-manus-founder-xiao-hong-the-andy-beal-law-of-the-new-era-2/ -
Everything We Know About the Founder of Manus AI So Far
https://hybrid-rituals.com/everything-we-know-about-the-founder-of-manus-ai-so-far/ -
Meta acquires Manus for US$2 billion, who is Xiao Hong?
https://www.idnfinancials.com/news/60049/meta-acquires-manus-for-us2-billion-who-is-xiao-hong -
Meet Xiao Hong: CEO of Manus, Chinese AI firm acquired by Mark Zuckerberg's Meta for $2 billion
https://www.financialexpress.com/life/technology-meet-xiao-hong-ceo-of-manus-chinese-ai-firm-bought-by-mark-zuckerbergs-meta-for-2-billion-4092361/ -
Assessing Manus: The Future of Agentic AI
https://www.infotech.com/research/assessing-manus-the-future-of-agentic-ai -
Manus Joins Meta for Next Era of Innovation
https://manus.im/blog/manus-joins-meta-for-next-era-of-innovation -
实探｜一夜爆火的Manus，背后团队是什么来头
https://m.bjnews.com.cn/detail/1741365583129817.html -
Overview of MANUS AI Agent. Company Background and Project Overview
https://medium.com/@astropomeai/overview-of-manus-ai-agent-6b1f37d90a91 -
Meta to buy Chinese founded startup Manus to boost advanced AI
https://www.reuters.com/world/china/meta-acquire-chinese-startup-manus-boost-advanced-ai-features-2025-12-29/