AI tools·Coding agents·China
Manus
Next-gen general autonomous agent. Excels at complex market research, data analysis, and multi-step web tasks.
Compiled through 2026-09
Invite-gated signup
- Vendor
- Monica Team
- Started
- 2025
- China availability
- Limited in ChinaReachable but constrained: may need your own API key, an invite, or unstable routing
- Pricing tier
- Free tier
- Current flagship
- Manus 1.5 (General Agent)
- Website
- manus.im
Evaluation, Pros & Cons
Editorial score based on public benchmarks and community feedback — not a third-party benchmark result
Key Advantages
- Truly general autonomous agent executing multi-step research, scraping, and deliverables across the web
- Advanced long-horizon planning with automated sandbox setup, debugging, and rich deliverables
Limitations & Caveats
- High compute requirements for long missions; invites and strict quotas apply
Overview
Incubated by the Monica team in early 2025, Manus captured worldwide attention by expanding agents from code-centric tasks to general digital workflows. Operating in virtual sandboxes, Manus browses the live web, writes Python scripts to clean data, creates charts, and packages ready-to-use reports.
Direction
Strengthening enterprise workspace data isolation, local desktop tool access, and an open agent marketplace.
Version history
Jun 2026
LatestManus 1.5
Multimodal interaction and collaborative multi-agent execution.
View details
New capabilities
- Manus 1.5 cross-device orchestration and multi-agent groups
- Enterprise workspace isolation and streaming PII anonymization
What got stronger
- 3x longer horizon execution with superior fault tolerance
- Native integrations with Feishu, DingTalk, Salesforce, and Notion
Version notes
Manus 1.5 focused on enterprise reliability and security, allowing teams to delegate tedious multi-system workflows to managed agent swarms.
Oct 2025
Manus 1.2
Batch file workflows and multi-source analytics.
Mar 2025
Manus launch
Viral launch demonstrating end-to-end multi-step task execution.
View details
New capabilities
- General autonomous task execution loop breaking code-only boundaries
- Multi-step web automation: search, scraping, form filling, and credentials
What got stronger
- Directly produces structured reports, spreadsheets, and interactive web charts
- Exceptional self-healing and planning resilience on open-ended assignments
Version notes
Manus's debut in spring 2025 gave the world a glimpse of real General Agents. Handling research, supplier pricing comparisons, and document drafting, it demonstrated tangible utility beyond coding.