upscale_image¶
Upscale an image by 2× or 4× using NovelAI’s dedicated upscaler. The upscaler is model-independent — it works on any source image, regardless of how it was produced.
Parameters¶
Name |
Type |
Default |
Notes |
|---|---|---|---|
|
|
required |
Base64-encoded PNG/JPEG to upscale. |
|
|
|
|
Returns¶
list[ContentBlock] — base64 Image of the upscaled image + saved-path text.
Behavior¶
The upscaler is a separate NovelAI endpoint — it does not run a diffusion loop, so it’s faster than re-generating at a higher resolution.
Anlas cost scales with source resolution and
factor. Useestimate_anlas_costbefore a big batch.The output preserves the input’s color profile and aspect ratio. You cannot change aspect ratio during upscale.
Example¶
await ctx.session.call_tool("upscale_image", {
"image": image_b64,
"factor": 4,
})
CLI¶
uv run python -m novelai_image_mcp upscale ./portrait.png --factor 4
Tips¶
Tip
When to upscale vs re-generate?
Upscale when you already have a near-perfect image and just need higher resolution.
Re-generate when you want to change composition, lighting, or details. Re-generating at 2× the original resolution often produces cleaner results than upscaling, but costs more Anlas.
Memory usage
A 4× upscale of a 1024×1024 image produces a 4096×4096 output (~64 MB
uncompressed PNG). Make sure your NOVELAI_OUTPUT_DIR volume has space.
See also¶
Account balance — check Anlas before a batch