Features
Generation Modes
- txt2img — Generate images from text prompts.
- img2img — Transform an existing image with a prompt and a denoise strength.
- inpaint — Redraw a selected area of an image while keeping the rest intact.
Custom Models
You can use your own models in addition to the built-in ones:
- CPU/GPU mode: Import any SD1.5 checkpoint directly in the app. The conversion happens on-device.
- NPU mode: SD1.5 and SDXL models must be converted on a host machine first. See the Conversion Guide.
You can also try pre-converted community NPU models from:
- xororz/sd-qnn — SD1.5 models
- xororz/sdxl-qnn — SDXL models
- Mr-J-369 — community-converted models
- YuuiKurata — community-converted models
LoRA Support
LoRA weights can be added to custom CPU models at import time. The weight is baked into the model.
Prompt Weights
Emphasize or de-emphasize words in a prompt using the same syntax as AUTOMATIC1111:
(masterpiece:1.5), (low quality:0.6), 1girlEmbeddings
Custom textual inversion embeddings (e.g. EasyNegative) are supported. The file must be in SafeTensors format. If your embedding is a .pt file, convert it to .safetensors before importing.
Upscalers
Built-in 4× upscalers:
- realesrgan_x4plus_anime_6b — anime-tuned
- 4x-UltraSharpV2_Lite — general purpose