- Who it helps
- Developers and researchers evaluating XR coding agents.
- Availability
- October 21 preprint, revised October 22; the abstract links code, which was not executed.
The test surface
Its 112 tasks cover features, repairs and application generation. Tests combine scripted and adaptive interaction with visual and programmatic checks.
Reported limits
The strongest tested configuration reaches 50.9% first-attempt success. That result applies to the authors’ task set and setup.
GenXR-Bench tests coding agents through interaction ↗HCI Index W43 Toolbox ↗
What you can do
- Read the primary evidence and stated availability before planning a trial. Open reference ↗
Where the evidence stops
Abstract and metadata checked; no headset trial, repository inspection or result reproduction.
Sources read. Sources checked October 30, 2026. No installation, model inference or end-to-end workflow was executed.
Sources & further reading
Paper · paper
GenXR-Bench tests coding agents through interaction ↗
Establishes the subject, discovery context or stated verification limit.
William Huang and coauthors · page checked 2026-10-30
Web · analysis
HCI Index W43 Toolbox ↗
Establishes the subject, discovery context or stated verification limit.
HCI Index · page checked 2026-10-30
Coverage note: No qualifying current independent attention sample; original X posts unavailable. Reported outcomes were not independently reproduced.
The record over time (1)
Dated source checks and published explanations. An observation date can differ from the original release date.
- · Added to Current
GenXR-Bench evaluates generated extended-reality applications through runtime interactions, including movement and sensing.