03 / Jan 13 – 19, 2025 · 10 RELEASES
AI News Week 3, 2025
Week 3 2025 AI releases.
Vision & Image (3)
Magic-1 For-1
DA-Group, PKU · Text-to-Image
A method for efficient one-to-one image generation and manipulation, enabling high-quality personalized image synthesis with minimal computational overhead.
Lumina Image 2.0
Comfy-Org · Text-to-Image
A 10.6GB image generation model repackaged for ComfyUI in safetensors format, representing an optimized version of Lumina's text-to-image diffusion architecture.
Lumina Image 2.0
ComfyUI Community · Text-to-Image
A diffusion model for image generation that uses Gemma 2 2B as its text encoder, integrated with ComfyUI workflow support.
Video & Animation (1)
Light-A-Video
Shanghai AI Laboratory et al. · Video Editing
Training-free video relighting method using Consistent Light Attention and Progressive Light Fusion to maintain temporal coherence and lighting consistency across frames without requiring video-specific training data.
Audio & Speech (1)
ComfyUI_Sonic
smthemex · Audio Processing
A ComfyUI node implementation for Sonic audio generation and processing, enabling audio synthesis workflows within the ComfyUI node-based interface.
Agents & Automation (2)
OmniParser
Microsoft · Web Agents
A screen parsing system that converts UI screenshots into structured, interactable elements for autonomous agents. Enables screen understanding and interaction capabilities for AI systems to navigate and manipulate user interfaces.
MedRAX
bowang-lab · Agent Framework
A medical AI system combining retrieval-augmented generation with domain-specific knowledge for improved clinical reasoning and medical question answering.
Developer Tools (3)
Flux Jupyter
camenduru · Inference Engine
A Jupyter notebook interface for running Flux image generation model, enabling text-to-image synthesis with interactive experimentation capabilities.
BEN2 ComfyUI
Prama LLC · Inference Engine
ComfyUI node implementation for BEN2, enabling integration of the BEN2 model into the ComfyUI visual node-based workflow system for image generation and processing tasks.
Awesome ComfyUI
Kakachia777 (Community) · AI Infrastructure
Curated collection of resources, nodes, and extensions for ComfyUI, a node-based UI for Stable Diffusion and other generative models. Aggregates community tools and workflows for image generation workflows.
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.