Astria Video

Astria Gallery is a prompt library created by the Astria team, mainly providing reference and inspiration for AI image and video generation.

Main Features and Characteristics:

1. Multi-model Support: Covers mainstream AI model branches including sd15, sdxl1, sd21, flux1, flux-pro1, and gpt-image-1.
2. Video and Animation Parameters: Supports complex dynamic control parameters such as translation_x/y/z, rotation_3d_x/y/z, strength_schedule, diffusion_cadence, seed, fov, etc., enabling keyframe animation and 3D perspective transformation.
3. ControlNet Support: Provides ControlNet options like Pose, Depth, and MLSD for precise image structure control.
4. Image Enhancement Features: Supports Super-Resolution, HiRes fix, Face Correct/Inpaint Faces, color_grading, etc.
5. LoRA and Prompt Weighting: Supports introducing external LoRA models (e.g., lora:3451857:1.0) and weighted prompts.
6. Video Generation Models: Integrates video generation models like seedance_480p and seedance_720p, supporting the --video parameter and --video_prompt for specific video generation instructions.

Typical Use Cases:

• Cross-temporal age progression of characters (e.g., growing from a toddler to an old man).
• Stylized portrait generation (e.g., Gustav Klimt painting style, cyberpunk style, western cowboy, etc.).
• Scene transformation and transition (e.g., smooth shifts from desert to jungle, from indoor to cosmic starry sky).
• Product showcase (e.g., dynamic display of cars in different environments and styles).
• Microscopic world and surreal scene rendering (e.g., macro photography of phosphorescent spores, infinity mirror rooms, etc.).

The page does not detail specific pricing models and fees, but includes a Pricing link, indicating a paid service nature.