Picx Real is a text-to-image model hosted on Hugging Face under the repository GraydientPlatformAPI/picx-real. It belongs to the class of Stable Diffusion models and generates images from text prompts using the Diffusers library.
The model is provided in Safetensors format with 0.9 billion parameters and uses F16 tensor type. It runs via a DiffusionPipeline that loads the checkpoint with bfloat16 precision and supports CUDA device mapping for accelerated inference on compatible hardware. Example code demonstrates loading the pipeline, supplying a prompt such as "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k", and producing an output image. The repository also references notebooks on Google Colab and Kaggle as well as local applications including Draw Things and DiffusionBee.
It carries an OpenRAIL license. The model card itself contains no additional descriptive content beyond the usage instructions and technical specifications. Downloads reached 72,752 in the most recent month recorded on the platform.
In the Image generation space, Picx Real takes a focused approach. It focuses on generating realistic images from text prompts using diffusion models. It is built as an open-source project for AI artists and developers. The project is open source (Open Source). Picx Real is available on the web, the command line, and API.
Behind Picx Real is Graydient Platform, and it first shipped in 2023. Key capabilities include text-to-Image, Stable Diffusion, and Realistic Generation. It exposes integrations via a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match