- Who it helps
- Researchers and creative developers evaluating programmatic video generation.
- Availability
- The paper was submitted October 20. Its abstract and HTML version are public. The linked project site was inaccessible during this check, so downloadable artifacts remain unverified.
What is tested
The authors describe 2,104 tasks across ten categories and 24 subtypes, with text, image and video inputs. Their comparison covers fifteen models and reports degradation as task complexity increases.
How to interpret a score
The study combines objective and subjective measures. A result on this collection does not establish reliability for an unseen animation pipeline, rendering environment or production brief.
What you can do
- Read a task definition and its scoring rules before using a reported result to choose a model for an animation workflow. Open reference ↗
Where the evidence stops
Paper claims were not independently reproduced. The demo, code and dataset availability could not be established from the inaccessible project site.
Sources read. Original sources read October 23, 2026 UTC. No external project, model inference or reported experiment was executed.
Sources & further reading
Paper · paper
Code2VideoCompass ↗
Read the October 20 abstract and paper introduction. The benchmark concerns generating executable video programs; artifacts linked from the project page could not be verified.
Code2VideoCompass authors · page checked 2026-10-23 · Published 2026-10-20
Paper · paper
Code2VideoCompass v1 paper ↗
Read task composition and evaluation description. No implementation, dataset download or model run was performed.
Code2VideoCompass authors · page checked 2026-10-23
Coverage note: No independent end-to-end reproduction or complete video demonstration was verified in this bounded sample.
The record over time (1)
Dated source checks and published explanations. An observation date can differ from the original release date.
- · Added to Current
Code2VideoCompass evaluates models that turn text or visual inputs into programs which produce videos. The paper makes completed video quality part of the evaluation, beyond whether code looks plausible.