Hello, creators and builders,
With the full Qwen3-TTS family now open-sourced this week, anyone can now design, clone, or fine-tune voices with studio-grade quality, in 10 languages, using just 0.6B parameters.
Let’s dive in.
👉 Subscribe to The Tongyi Weekly and never miss a release:
Subscribe Now → https://www.linkedin.com/build-relation/newsletter-follow?entityUrn=7392460924453945345
📣 Model Release & Updates
Qwen3-TTS Full Model Family Open-Sourced
Qwen3-TTS is officially live. We’ve open-sourced the full family—VoiceDesign, CustomVoice, and Base—bringing high quality to the open community.
- 5 models (0.6B & 1.7B)
- Free-form voice design & cloning
- Support for 10 languages
- SOTA 12Hz tokenizer for high compression
- Full fine-tuning support
SOTA performance
Everything is out now—weights, code, and paper. Enjoy.
🧩 Ecosystem Highlights
Wan 2.6 Reference-to-Video Available in ComfyUI
Great news for visual creators: Wan 2.6 Reference-to-Video is now officially integrated into ComfyUI. Now you can turn 1–2 reference clips into short videos with natural motion, camera movement, and visual style.
- Learn motion, camera angles, and style from reference videos
- Combine up to two clips for blended motion and visual style
- Generate 5–10 second videos at 720p or 1080p
- Maintain character consistency and natural movement
✨ Community Spotlights
Refine Lighting Effects: Qwen-Edit-2511_LightingRemap_Alpha0.2 from zooeyy
This LoRA by zooeyy is specifically designed for "color-block-guided relighting." The LoRA intelligently reconstructs lighting and renders atmosphere on subjects or scenes based on the position and hue of color blocks in the input image, while automatically removing the blocks themselves to produce natural, high-fidelity illumination.
👉 Try it here
Object Remover: Qwen-Image-Edit-2511-Object-Remover from prithivMLmods
Qwen-Image-Edit-2511-Object-Remover by prithivMLmods is an adapter (LoRA) developed for Qwen’s Qwen-Image-Edit-2511 image-to-image model, specifically designed for precise object removal from images. The model removes specified objects while preserving the background and remaining elements, maintaining realism and original visual details.
👉 Try it here
Image Style Customization Workflow: AmazingZImageWorkflow from martin-rizzo
This workflow for Z-Image-Turbo by martin-rizzo expands the ComfyUI base workflow with additional features, particularly focused on high-quality image styles and user-friendly functionality, while also integrating an image refiner and a simple upscaler. Its standout feature is a dedicated Style Selector with 18 customizable presets, allowing you to switch artistic directions instantly.
👉 Try it here
Z-Image Nodes Collection: ComfyUI-ZImagePowerNodes from martin-rizzo
This toolkit by martin-rizzo solves common pain points by offering one-click style switching and high-speed "Turbo" configurations (optimizing for as few as 7 steps).
It effectively brings precision and ease-of-use to the Z-Image ecosystem.
👉 Try it here
📬 Want More? Stay Updated.
Every week, we bring you:
- New model releases & upgrades
- AI research breakthroughs
- Open-source tools you can use today
- Community highlights that inspire
👉 Subscribe to The Tongyi Weekly and never miss a release.
Subscribe Now → https://www.linkedin.com/build-relation/newsletter-follow?entityUrn=7392460924453945345
Thank you for being part of this journey.
Tongyi Lab is a research institution under Alibaba Group dedicated to artificial intelligence and foundation models, focusing on the research, development, and innovative applications of AI models across diverse domains. Its research spans large language models (LLMs), multimodal understanding and generation, visual AIGC, speech technologies, and more.
Top comments (0)