ComfyUI Extension

SpicyAPI

Every SpicyAPI image, video and audio model as a ComfyUI node: Seedance, Kling, Wan, MiniMax, Suno, GPT Image, Nano Banana and 180+ more, billed per request in USD.

By spicyapi-ai·Created a day ago·Updated about 23 hours ago· 0

spicyapi-ai/comfyui-spicyapi

Nodes—

On cloudLocal install

Stars0

Updatedabout 23 hours ago

Readme

SpicyAPI for ComfyUI

Seedance, Kling, Wan, MiniMax H3, Suno, GPT Image, Nano Banana and more AI models as ComfyUI nodes. No GPU, no model downloads.

Get an API key · Setup guide · All models · Status

</div>

comfyui-spicyapi is a ComfyUI custom node pack that runs hosted SpicyAPI models from inside your workflows. Each model is one node whose options come from that model's published input schema, so the node offers exactly what the model accepts. Generation runs on SpicyAPI's servers: you need an internet connection and an API key, not a graphics card or model weights. Every request is paid in US dollars from a prepaid balance, and each node shows its price before it runs.

A ComfyUI graph built from SpicyAPI nodes: Z-Image Spicy Pro text to image feeding Wan 2.2 Spicy image to video, each node showing its price badge and the final charge

189 model nodes (104 video, 69 image, 16 audio) from 117 model families, plus 25 text models in one Chat node.

At a glance

| | | | --- | --- | | What it adds | One node per hosted model, a Chat node for text models, and a LoRA node | | Inputs and outputs | Standard ComfyUI IMAGE, MASK, VIDEO, AUDIO and STRING | | Hardware | None: models run on SpicyAPI, so it works on laptops, Macs and CPU-only machines | | Dependencies | None beyond ComfyUI itself (Python 3.10+, the V3 node API; tested with ComfyUI 0.39) | | Pricing | Pay per request from a prepaid USD balance; the rate is shown on every node | | New models | Picked up automatically: the node list is rebuilt from your account at each start | | License | MIT |

Install

cd ComfyUI/custom_nodes
git clone https://github.com/spicyapi-ai/comfyui-spicyapi

Then restart ComfyUI. Prefer not to use git? Download the ZIP, unpack it into ComfyUI/custom_nodes, and restart. There is nothing else to install.

Quick start

Add your key. Open Settings > SpicyAPI, paste a key from spicyapi.ai/console/keys; a notice shows your balance. Restart ComfyUI once so the node list comes from your account.
Add a model node. Double-click the canvas and type a model name (seedance, kling, wan, suno, nano banana...), or browse SpicyAPI > Video / Image / Audio in the node menu.
Connect and run. Wire a Load Image into the node's image input if the model takes one, connect Save Image, Save Video or Save Audio to its output, and press Run. The node shows its progress, then what the run cost.

Prefer to start from a working graph? Open ComfyUI's template browser: under comfyui-spicyapi you will find text to image, image to video, a chat model writing the prompt for an image model, text to speech, and LoRA.

Supported models

The node list follows the live SpicyAPI catalogue. At the time of this release:

Text (25 models in the SpicyAPI Chat node)

Claude Sonnet 5.5, Claude Opus 5.5, Grok 4.7, DeepSeek V4.1 Flash, GLM 5.3 Flash, GLM 5.3, DeepSeek V4 Pro, Gemini 3.7 Flash, Grok 4.6, DeepSeek V4 Flash, Claude Opus 5, Gemini 3.5 Flash-Lite, Kimi K3, Grok 4.5, Claude Sonnet 5, GLM 5.2, Claude Fable 5, Grok 4.3, Gemini 3.1 Pro Preview, Claude Sonnet 4.6, Claude Opus 4.6, Gemini 3 Flash Preview, Claude Opus 4.5, Claude Sonnet 4.5, Gemini 2.5 Flash.

Every node with its model ID and starting price: MODELS.md.

How the nodes work

Inputs follow the model. Dropdowns, sliders and switches mirror each model's schema. Image, video and audio inputs are sockets, uploaded for you when the node runs. Models that take several references grow a new socket every time you connect one.
Defaults are the model's own. A dropdown set to model default, or a number set to -1, is left out of the request so the model decides.
Outputs are ordinary ComfyUI values. Video nodes return VIDEO (and the last frame as IMAGE when return_last_frame is on), audio nodes return AUDIO, image nodes return IMAGE, and background removal also returns an alpha mask. Transcription returns STRING.
SpicyAPI Chat sends a prompt, optional system instructions and optional images to any text model on SpicyAPI and returns the answer as a STRING, ready to wire into another node's prompt.
SpicyAPI LoRA adds a LoRA weights file by URL to models with LoRA support. Chain several to stack them.

A SpicyAPI LoRA node with a pixel-art LoRA link feeding Z-Image Turbo LoRA, and the pixel-art result

Costs and safety

Price on the node. Each node's badge shows the current rate for the options you picked, for example $0.0525/s for video or $0.01/image, and updates as you change them.
Quote before every run. Set Settings > SpicyAPI > Max cost per run and any run quoted above it stops before anything is charged.
No surprise re-runs. Seeds start on fixed, so queueing the same graph twice reuses the first result. Switch control after generate to randomize for a new result each run.
Your key stays out of shared files. It is stored in the ComfyUI user folder (user/spicyapi/config.json, readable by your account only), never in a workflow or the metadata of a saved image. On a server, set SPICY_API_KEY instead.
Cancelling stops the wait, not the job. A cancelled run keeps going on SpicyAPI, is charged if it succeeds, and its result is in your console for up to 14 days.

FAQ

How do I use Seedance in ComfyUI?

Install this pack and add your key, then add the Seedance 2.5 · Image to Video node (or any other Seedance node), connect Load Image to its image input and Save Video to its video output, write a prompt and run. Resolution, duration and audio are options on the node. The same steps work for Kling, Wan, MiniMax H3, Vidu and LTX video models.

Do I need a GPU or to download model weights?

No. The models run on SpicyAPI's servers, and the nodes only send inputs and receive results. A laptop or a CPU-only machine is enough.

How much does a run cost?

Each model has its own rate, shown on the node and on its page at spicyapi.ai/models: image models are priced per image, video models per second of output, speech per character, per second or per request. A run is quoted before it starts, and the node prints the final charge when it finishes. Failed runs are not charged unless a model's page says otherwise.

Is my API key safe when I share a workflow or an image?

Yes. ComfyUI writes node settings into workflow files and into PNG metadata, so the key is never a node setting. It lives in a file in your ComfyUI user folder, or in the SPICY_API_KEY environment variable.

Can I use my own LoRA?

Yes, with the LoRA model nodes (their names end in LoRA). Add a SpicyAPI LoRA node with a direct download link to a .safetensors file and its strength, and connect it to the model node's loras input. The link must open without a login. The lora template comes ready to run with an Apache-2.0 pixel-art LoRA for Z-Image Turbo (by tarn59 on Hugging Face); swap in your own link.

A model I saw on spicyapi.ai is not in my node list

Restart ComfyUI. The node list is built from your account's catalogue when ComfyUI starts, so models added since the last start appear after a restart. Without a key, the list bundled with this release is used.

Does it work with ComfyUI on a remote server or in the cloud?

Yes, as long as that machine can reach api.spicyapi.ai. Set SPICY_API_KEY in its environment, or enter the key in Settings from the browser. Proxies set through HTTPS_PROXY are honoured.

Troubleshooting

| You see | What to do | | --- | --- | | No SpicyAPI key is set | Add the key in Settings > SpicyAPI, or set SPICY_API_KEY. | | 40310 email not verified | Open the verification link from your sign-up email. | | 40004 no deployment serves this combination | Change the option named in the message. | | An error ending in request_id ... | Send that id to support; it identifies the request exactly. | | a proxy or firewall ... may be intercepting the request | Check that HTTPS_PROXY, if set, points at a working proxy. |

Other ways to use SpicyAPI

Python SDK, TypeScript SDK, Go, PHP and Java clients
CLI for the terminal, and an MCP server for AI assistants
Documentation and the API reference

Development

python -m unittest discover -s tests -t .
python scripts/update_snapshot.py   # needs SPICY_API_KEY; refreshes the bundled model list
python scripts/render_docs.py       # rewrites the model lists in README.md and MODELS.md

License

MIT