Manus 15-task agent test
The JustPickAi editorial team gave Manus 15 real-world tasks across research, data, coding and content, then timed and scored each one.
In the published table, most tasks succeeded in 4 to 15 minutes, with 70% full success, 20% partial and 10% failure.
Overview
How it was set up
Manus run in its cloud-based virtual environment, with no extra tools or integrations stated.
Worth knowing
Technical tasks did poorly: Python debugging was partial and a multi-step API integration failed. The review also warns against using it for sensitive data.
A prompt to start from
The source does not show the prompt.
Sources 1
History
- Source published
- 2026-02-20
- Added to Agentic Feats
- 2026-10-13
- Pick
- No
- Markdown copy
- /c/manus-15-task-agent-test.md
More in research
36 featsResultIt created the text file as asked, but after clicking a result it took a screenshot too early and hallucinated the answer.
ResultOperator handled tour search and lesson plan but was slow; adding three grocery items took 15 minutes and an Uber price check took five.
ResultThe draft was usable but weaker than accepted NSF summaries; thinking mode helped, though it began hallucinating and misreading the sources.