Chroma-GGUF provides multiple quantized GGUF versions (from 2-bit to 8-bit) of a 9B parameter text-to-image model based on Flux architecture. The files are hosted on Hugging Face and are compatible with GGUF-supporting inference engines. It enables running high-quality image generation locally on a wide range of hardware.
In the Image generation space, Chroma takes a focused approach. It focuses on making large text-to-image models runnable on consumer hardware through GGUF quantization. It is built as an open-source project for Local AI inference users. The project is open source (Open Source). Chroma is available on the web and API.
Behind Chroma is silveroxides, and it first shipped in 2024. Among its 3 catalogued features are GGUF quantization, multiple bit widths, and text-to-image.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do