Rax-4.5 is a large multimodal model supporting text, image, and video inputs. It features a sophisticated chat template for handling mixed media content. The model is hosted on Hugging Face and is intended for local or self-hosted inference environments.
In the Foundation models & chat space, Rax takes a focused approach. It focuses on processing and reasoning over images, video, and text in a single unified model. Rax is an open-source project aimed at developers. Rax is open source under the Apache-2.0 license. Rax is available on the web, the command line, and API.
raxcore-dev builds and maintains Rax, and it first shipped in 2023. The project is developed in the open on GitHub with 30.6k stars and 3.6k commits in the last 90 days. Among its 3 catalogued features are Vision Understanding, Video Processing, and Multimodal Chat.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Other projects tracked alongside this one