ComfyUI Node
🤖 Leon LLM JSON API
A ComfyUI node in Leon_API with 11 inputs and 1 output.
🤖 Leon LLM JSON API
- input_image
- image_array
- json_response
◄model►
◄json_schema{"type": "object", "properties": {"email": {"type": "string", "description": "The extracted email address"}}, "required": ["email"]}►
◄api_urlhttps://api.hyprlab.io/v1/chat/completions►
◄api_keyYOUR_HYPRLAB_API_KEY►
◄system_messageYou are a helpful assistant that extracts structured data.►
◄max_tokens128000►
◄temperature0.1►
◄image_url►
CategoryLeon_API
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| model | STRING | Model name to use for JSON completion | |
| user_message | STRING | Extract the email address from this text: Contact us at [email protected] | User message to send to the model |
| json_schema | STRING | {"type": "object", "properties": {"email": {"type": "string", "description": "The extracted email address"}}, "required": ["email"]} | JSON schema for structured output |
| api_url | STRING | https://api.hyprlab.io/v1/chat/completions | API URL for chat completions |
| api_key | STRING | YOUR_HYPRLAB_API_KEY | Your HyprLab API key |
| system_messageopt | STRING | You are a helpful assistant that extracts structured data. | System message to set the assistant's behavior |
| max_tokensopt | INT | 1280001–1048576 | Maximum number of tokens to generate |
| temperatureopt | FLOAT | 0.10–2 | Sampling temperature (lower for more consistent JSON) |
| input_imageopt | IMAGE | Optional single image input for vision-capable models | |
| image_urlopt | STRING | Optional image URL for vision-capable models | |
| image_arrayopt | IMAGE_ARRAY | Optional array of images (base64 or URLs) for vision-capable models |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| json_response | STRING | — |