OmniParser ‘tokenizes’ UI screenshots from pixel spaces into structured elements in the screenshot that are interpretable by LLMs. This enables the LLMs to do retrieval based next action prediction given a set of parsed interactable elements.
Turn any LLM into a Computer Use Agent
OmniParser V2
Subscribe to NBW newsletter
CC
An AI agent built for families
Google Earth AI
Combines health data with geospatial insights to predict and reduce health risks.
Falcon-Emirati-7B
A dialect-specialized model for understanding and generating Emirati Arabic.
Google Flow Music
Producers can create custom music production tools with ease.
EmbeddingGemma 2
A lightweight model for on-device multimodal embeddings.
Personal Agent Protocol
An open standard for personal agents to interact with businesses.