Flux has long been regarded as the state-of-the-art in image generation. But there’s a new contender making waves—WAN—delivering even more impressive results, especially for realistic, cinematic shots.
👉 Try it now: https://textideo.com/model/wan2-2
Introduction
While Flux continues to be widely praised, WAN, originally designed for video generation, has revealed a surprising strength in single-frame image creation.
By simply setting the frame count to 1, WAN transforms into a highly capable image generator trained on cinematic visuals, producing shots with stunning detail and atmosphere.
Get 100 Free Credits
New users registering on Textideo
will receive 100 free credits instantly.
These credits let you explore all features of the model right away—no upfront commitment required. It’s a great way to test the capabilities and see what you can create before subscribing.
What You’ll Get
WAN uses a text-to-image workflow enhanced with several LoRAs for optimized output:
Detailz-WAN → Enhances facial details, skin texture, depth, color, and lighting.
MPS Rewards → Reward backpropagation improves prompt adherence and overall quality.
LightX2V Fast → Speeds up generation for faster results.
FusionX → Balances speed, prompt alignment, and fluidity in the outputs.
👉 Workflow & demo: https://textideo.com/model/wan2-2
The LoRAs come pre-tuned for optimal results, while still allowing for customization.
What It Excels At
WAN is ideal for cinematic, high-impact visuals. One of the most practical applications is creating ad posters, especially when combined with product imagery or other visual enhancement techniques.
Getting Started
Running the workflow is simple:
On Textideo → Click the Launch Workflow button and enter your prompt.
Locally → Download the workflow and models, then follow the setup instructions below.
Models
For local usage:
Download the workflow file.
Access all models via the provided links on the workflow page.
Nodes
To fully customize, download the following packs:
ComfyUI Impact Pack
ComfyUI-GGUF
ComfyUI-KJNodes
ComfyUI-Crystools
VRGameDevGirl Video Enhancement Nodes
ComfyUI-GGUF-FantasyTalking
Step-by-Step Guide
On Textideo, enter your prompt in the Positive Prompt node.
Click Run.
Locally, ensure:
Models are in the correct directories.
LoRA and model nodes are linked correctly.
The EmptyHunyuanLatentVideo node is set to length = 1.
Examples
Community
Have some great shots from WAN? 🎬
Join our Discord to share your creations and get inspired by the community!
👉 Explore WAN today: https://textideo.com/model/wan2-2
Top comments (0)