VISTA
A research harness for playing ARC-AGI-3, often used as the comparison point for other agent systems.
VISTA is a harness, meaning the scaffolding built around a language model so it can interact with an environment rather than just answer questions. It was designed for ARC-AGI-3, the benchmark where an agent is dropped into an unfamiliar game with no rules and has to work everything out by poking at it. VISTA shows the model a rendered picture of the game board and lets it act directly, without first building an internal program that models how the world works.
Because it was one of the first systems to solve the full public set, VISTA has become the reference other teams measure against. When you see a claim like “12 percent fewer actions”, it is usually VISTA on the other side of the comparison.