- Who it helps
- Developers evaluating browser-agent interfaces.
- Availability
- Public MIT-licensed 0.1.x beta; repository documentation checked November 6.
Where it helps
The interface aims to reduce the text an agent reads before acting. It cannot replace visual interaction with canvas drawings, maps or WebGL applications.
Read the comparison carefully
The maintainer reports lower token use on a small task set. Two iframe/tab tasks were excluded from correctness scoring, some tasks were selected after results, and the screenshot baseline is the maintainer’s implementation.
What you can do
- Read the linked primary evidence and its access limits before planning a trial. Open reference ↗
Where the evidence stops
No browser task or benchmark was run. Login and bot walls may require human interaction; the linked demo was inaccessible.
Sources read. Sources read November 6, 2026 UTC. No third-party project, model inference or complete product workflow was executed.
Sources & further reading
GitHub · official
Mantis presents a compact text view of a browser ↗
Primary page read November 6. Claims were checked against the page; the workflow was not executed.
perzianhq / perzian · page checked 2026-11-06
Coverage note: Original YouTube demonstrations and independent cross-platform attention were not established. No end-to-end workflow was executed.
The record over time (1)
Dated source checks and published explanations. An observation date can differ from the original release date.
- · Added to Current
Mantis gives an agent numbered browser controls in a short text view.