Janus Pro: Multimodal Image Understanding and Generation for ComfyUI

Janus Pro is DeepSeek's multimodal AI model for image understanding and generation, available in 1B and 7B variants.

J

Janus Pro

MultimodalImage UnderstandingText-to-ImageOpen Source

DeepSeek's multimodal AI model open-sourced on January 27, 2025. Capable of both image understanding and generation. Available in 1B and 7B parameter variants. MIT licensed. Supports image captioning, visual question answering, and text-to-image generation via ComfyUI-Janus-Pro plugin.

Hugging Face (7B)Hugging Face (1B)GitHub
DeveloperDeepSeek
Release Date2025-01-27
ArchitectureMultimodal Transformer (autoregressive)
LicenseMIT
Model Sizes7B (recommended), 1B (lightweight)
CapabilitiesImage understanding, visual QA, text-to-image generation
PluginComfyUI-Janus-Pro (CY-CHENYUE)

Guides and workflows related to this model series.

No articles found.