You're looking at a specific version of this model. Jump to the model overview.
philz1337x /crystal-upscaler:64460f86
Input schema
The fields you can use to run this model with an API. If you don’t give a value for a field its default value will be used.
| Field | Type | Default value | Description |
|---|---|---|---|
| image |
string
|
An input image for video upscaling
|
|
| new_resolution |
integer
|
1024
|
Target resolution for upscaling (lower values use less VRAM)
|
| batch_size |
integer
|
1
|
Batch size for processing (1 recommended for most cases)
|
| preserve_vram |
boolean
|
True
|
Enable VRAM preservation (recommended for limited GPU memory)
|
| blocks_to_swap |
integer
|
32
|
Number of blocks to swap for memory optimization (higher = more memory efficient)
|
| use_non_blocking |
boolean
|
True
|
Use non-blocking operations for better performance
|
| offload_io_components |
boolean
|
True
|
Offload I/O components to save VRAM
|
| cache_model |
boolean
|
True
|
Cache model in memory (faster but uses more VRAM)
|
| enable_debug |
boolean
|
False
|
Enable debug mode for troubleshooting
|
| tiled_vae |
boolean
|
True
|
Process VAE in tiles to reduce VRAM usage (enable if running out of memory)
|
| vae_tile_size |
integer
|
512
Min: 64 Max: 2048 |
VAE tile size in pixels (smaller = less VRAM but more artifacts)
|
| vae_tile_overlap |
integer
|
64
Min: 16 Max: 256 |
VAE tile overlap in pixels (higher = fewer seams but slower)
|
| keep_vae_loaded_during_decode |
boolean
|
False
|
Keep VAE loaded in memory during decode (faster but uses more VRAM)
|
| output_format |
None
|
png
|
Format of the output images
|
| output_quality |
integer
|
100
Max: 100 |
Quality of the output images, from 0 to 100. 100 is best quality, 0 is lowest quality.
|
| seed |
integer
|
Set a seed for reproducibility. Random by default.
|
Output schema
The shape of the response you’ll get when you run this model with an API.
Schema
{'items': {'format': 'uri', 'type': 'string'},
'title': 'Output',
'type': 'array'}