Pass IndexThe State of AISign in

Vision-language model

A model that reads images alongside text — screenshots, documents, charts, photographs.

how it works · the vocabulary

VLMmultimodal modelvision

Images are converted into tokens, so a high-resolution page can cost as much as several thousand words. Most document processing that used to need OCR and a layout engine is now one call to one of these.

In the catalogue

Nearby

AgentAgent memoryAlignmentAutoregressiveBM25ChunkingCold startComputer use