Manus
Autonomous AI agent executing complex multi-step tasks from plain-language objectives.
1. Core Product / Service
Manus is a general-purpose autonomous AI agent that accepts an objective in natural language and independently plans and executes tasks without human intervention. Launched March 6, 2025, in closed beta, it operates in a cloud environment and can complete tasks including resume screening, stock analysis, trip planning, website creation, and real-estate research.
The product integrates US-based infrastructure built on Claude Sonnet, Microsoft Azure, and open-source browser automation tools. A key differentiator is the March 2026 "My Computer" feature—a desktop app (macOS/Windows) enabling agents to operate directly on user machines, expanding task scope beyond browser-based workflows.
Manus 1.5 (October 2025) reduced average task completion time from 15 minutes to under 4 minutes. By June 2026, the platform claimed $450 million estimated annualized revenue, up from $127 million earlier that year, with monthly growth exceeding 20%.
2. Target Users & Pain Points
Consumer: Individual users seeking AI assistance with knowledge work, research, scheduling, and administrative tasks. Freemium entry at $39/month (3,900 credits, 2 concurrent tasks) captures price-sensitive segment; waitlist exceeded 500,000 by mid-2026.
Enterprise: Organizations automating customer service, internal workflows, and data analysis. Microsoft partnership positions Manus as a launch partner for Microsoft Agent 365, converting platform dependency risk into distribution channel.
Pain point: Existing AI assistants (ChatGPT, Gemini) require manual step-by-step guidance; Manus removes need for human oversight on defined tasks.
3. Competitive Landscape
| Competitor | Positioning | Differentiation |
|---|---|---|
| OpenAI (ChatGPT agents) | Bundled into broader platform | Manus: vendor-neutral, standalone |
| Cognition (Devin) | Engineering-focused code agent | Manus: horizontal, multi-domain |
| Adept | Raised as specialized agent | Manus: consumer-first, rapid iteration |
| Genspark | AI search + workspace | Manus: execution-focused, not research |
| Writer (enterprise agents) | Vertical SaaS | Manus: horizontal + vertical SDKs |
Manus avoids lock-in by remaining platform-agnostic and offering enterprise SDKs for specialized verticals (finance, HR, legal).
4. Unique Observations
Regulatory arbitrage failed. Butterfly Effect (Manus' parent) was founded in China, relocated to Singapore in mid-2025 after laying off China-based staff. Despite geographic hedging, China's NDRC blocked Meta's $2B+ acquisition on April 27, 2026, citing national security. Meta withdrew June 15, 2026—a rare case of a Chinese-founded startup escaping US acquisition precisely because US-China tensions now make foreign ownership of AI infrastructure a strategic chokepoint.
Revenue velocity is genuinely fast. $450M ARR within 15 months of launch (March 2025 → June 2026) rivals post-launch growth of genspark (search) and early elevenlabs (voice). The invite-only scarcity and invitation code resales (¥50–100k in secondary markets) suggest genuine demand compression, not hype inflation.
Desktop agent plays a different game. "My Computer" feature (March 2026) and Microsoft SDK partnership hint at shift from cloud-only agent to OS-level automation layer—if executed at scale, this is a Zapier → UiPath competitive repositioning, not just GenAI consumer toy.
5. Financials / Funding
- Total raised (primary equity): $0.16B
- Latest valuation: $2.0B
| Date | Round | Amount | Post-money | Lead investor(s) |
|---|---|---|---|---|
| 2023-02 | Seed | undiscl. | $0.0B | ZhenFund |
| 2023-08 | Angel | undiscl. | $0.1B | ZhenFund |
| 2024-11 | Series A | $0.01B | $0.1B | HSG (HongShan / Sequoia China) |
| 2025-04 | Series B | $0.07B | $0.5B | Benchmark |
| 2025-12 | Strategic / Acquisition (M&A, not primary capital) | $2.00B | $2.0B | Meta Platforms |
| 2026 | Tender / Secondary buyback (not primary capital) | $2.00B | $2.0B | Tencent |
Note: Meta acquisition (Dec 2025, announced ~$2–3B) blocked by Chinese NDRC (April 27, 2026); deal formally terminated June 15, 2026. Valuation includes last Series B ($500M post-money, April 2025); subsequent Meta deal was structurally aborted and did not close.
6. People & Relationships
Founders:
- Xiao Hong – Founder of Butterfly Effect (2022); founded Nightingale Technology (2015); also launched Monica (AI browser extension).
- Ji Yichao ("Peak") – Chief scientist; founder of Peak Labs; created Mammoth Browser (iOS app) in high school.
- Zhang Tao – Product director.
Investors:
- Benchmark (Series B lead, April 2025; $75M at $500M post)
- Sequoia China / HongShan (Series A, Nov 2024)
- ZhenFund (Seed & Angel, 2023)
Strategic Partners:
- Microsoft – Agent 365 launch partner; enterprise SDK distribution
- Anthropic (Claude Sonnet backbone)
- Meta Platforms (blocked acquisition, June 2026 withdrawal)
Secondary investors: Tencent (reported secondary interest, 2026).