DeepSeek Janus: Multimodal AI Model for ComfyUI

DeepSeek Janus Pro is a multimodal AI model capable of both image understanding and generation, open-sourced by DeepSeek in January 2025.

D

DeepSeek Janus Pro

MultimodalImage UnderstandingText-to-ImageOpen Source

Multimodal AI model by DeepSeek capable of both image understanding and generation. Available in 1B and 7B variants. MIT licensed. Supports image captioning, visual question answering, and text-to-image generation.

Hugging Face (7B)GitHubPaper
DeveloperDeepSeek
Release Date2025-01-27
ArchitectureMultimodal Transformer (autoregressive)
LicenseMIT
Model SizesJanus-Pro-7B, Janus-Pro-1B, Janus-1.3B, JanusFlow-1.3B
CapabilitiesImage understanding, visual QA, text-to-image generation