MuAPI Example Workflows
Drop-in *.json workflows for the MuAPI ComfyUI node pack. Open ComfyUI → drag any
.json file onto the canvas (or use Workflow → Load).
API key: every node has an empty api_key field. Run muapi auth configure --api-key YOUR_KEY
once and all workflows pick up the key from ~/.muapi/config.json. Or paste your key into
any one node's api_key field per workflow.
Text → Image
| Workflow |
Model |
MuAPI_Basic_TextToImage_Flux.json |
flux-dev-image (great default) |
MuAPI_TextToImage_Imagen4.json |
google-imagen4-ultra |
MuAPI_TextToImage_Seedream5.json |
seedream-5.0 |
MuAPI_TextToImage_GPT_Image.json |
gpt-image-1.5 |
MuAPI_TextToImage_HiDream.json |
hidream_i1_full_image |
Image → Image (edit)
| Workflow |
Model |
MuAPI_ImageToImage_FluxKontext.json |
flux-kontext-pro-i2i |
MuAPI_ImageToImage_Seededit.json |
bytedance-seededit-image |
MuAPI_ImageToImage_GPT4o_Edit.json |
gpt4o-edit |
Text → Video
| Workflow |
Model |
MuAPI_TextToVideo_Seedance.json |
seedance-v2.0-t2v |
MuAPI_TextToVideo_Veo3.json |
veo3.1-text-to-video |
MuAPI_TextToVideo_Kling.json |
kling-v2.6-pro-t2v |
Wiring video into the saver: the MuAPIVideoSaver node takes video_url as a
string widget. After loading a T2V workflow, right-click the saver's video_url
field → Convert to input, then drag a link from the T2V node's video_url
output to the new input slot. (Some workflows already have this hooked up — see
the Chain_* examples.)
Image → Video
| Workflow |
Notes |
MuAPI_ImageToVideo_Seedance.json |
LoadImage → Seedance i2v |
MuAPI_ImageToVideo_Veo3.json |
LoadImage → Veo 3.1 i2v |
MuAPI_ImageToVideo_4References.json |
4 reference images → Seedance Omni |
Video extend / edit
| Workflow |
Notes |
MuAPI_ExtendVideo.json |
Generate then extend the same clip |
MuAPI_VideoEdit_Effects.json |
video-effects style transfer |
MuAPI_VideoUpscale_Topaz.json |
Topaz Labs video upscale |
MuAPI_Lipsync_Sync.json |
Sync.ai lipsync |
Image enhance
| Workflow |
Model |
MuAPI_FaceSwap.json |
ai-image-face-swap (2 inputs: base + face) |
MuAPI_Upscale_Topaz.json |
topaz-image-upscale (4×) |
MuAPI_BackgroundRemover.json |
ai-background-remover |
MuAPI_Ghibli_Style.json |
ai-ghibli-style |
MuAPI_ProductShot.json |
ai-product-shot |
Audio
| Workflow |
Model |
MuAPI_Audio_Suno.json |
suno-create-music |
Chains (multi-step pipelines)
| Workflow |
Pipeline |
MuAPI_Chain_T2I_to_I2V.json |
Flux 2 Pro → Veo 3.1 i2v |
MuAPI_Chain_I2I_then_Upscale.json |
Flux Kontext edit → Topaz 2× upscale |
Generic / advanced
| Workflow |
Notes |
MuAPI_Generic_RawJSON.json |
Call any endpoint with raw JSON + __file_N__ placeholders |
MuAPI_ApiKey_Hub.json |
API Key node wired to a T2I — paste your key once |
Regenerating
The _generate_workflows.py script regenerates every .json from a single spec
table. Edit it to add new examples and run:
Anil Matcha
41e3404a73