Use GenAI Models

These actions provide ability to use GenAI models from various providers.

Use GenAI Models Actions

LLMChat

Performs a chat interaction with a target LLM model provider. Event metadata fields applicable for this action are as follows:

Fields (table)

Field	Definition	Example	Default
Domain	ID of the model to use	chatgpt_chat	-
Input Element	Json path for the input in request event payload	message	-
Output Element	Json path for the output in response event payload	output	-

Fields (JSON Schema)

Input element can include the following fields:

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "chat": {
      "type": "string",
      "description": "ID of an ongoing chat for an assistant type use case with memory",
      "examples": ["test-chat"],
      "default": null
    },
    "message": {
      "description": "A text message or an object representing complex chat contents",
      "default": null,
      "oneOf": [
        {
          "type": "string",
          "examples": ["Hello"]
        },
        {
          "type": "object",
          "examples": [
            {
              "text": "Can you describe this image",
              "imageBase64": "base64"
            }
          ],
          "properties": {
            "text": {
              "type": "string",
              "description": "Text part of the user message",
              "examples": ["Hello"],
              "default": null
            },
            "textFile": {
              "type": "string",
              "description": "URI of a user message stored as a file",
              "examples": ["https://example.com/message.txt"],
              "default": null
            },
            "image": {
              "type": "string",
              "description": "URI of an image attachment, can be URL or data URI",
              "examples": ["data:image/png;base64,iVBORw0KGgoAAAANSUhE..."],
              "default": null
            },
            "imageBase64": {
              "type": "string",
              "description": "Base64 data of an image attachment",
              "examples": ["iVBORw0KGgoAAAANSUhE..."],
              "default": null
            },
            "imageBase64Mime": {
              "type": "string",
              "description": "Mime type of base64 image",
              "examples": ["image/jpg"],
              "default": null
            },
            "pdf": {
              "type": "string",
              "description": "URI of a PDF attachment",
              "examples": ["https://example.com/file.pdf"],
              "default": null
            },
            "pdfBase64": {
              "type": "string",
              "description": "Base64 data of a PDF attachment",
              "examples": ["iVBORw0KGgoAAAANSUhE..."],
              "default": null
            },
            "audio": {
              "type": "string",
              "description": "URI of an audio attachment",
              "examples": ["https://example.com/audio.mp3"],
              "default": null
            },
            "audioBase64": {
              "type": "string",
              "description": "Base64 data of an audio attachment",
              "examples": ["iVBORw0KGgoAAAANSUhE..."],
              "default": null
            },
            "video": {
              "type": "string",
              "description": "URI of a video attachment",
              "examples": ["https://example.com/video.mp4"],
              "default": null
            },
            "videoBase64": {
              "type": "string",
              "description": "Base64 data of a video attachment",
              "examples": ["iVBORw0KGgoAAAANSUhE..."],
              "default": null
            }
          }
        }
      ]
    },
    "messages": {
      "type": "array",
      "description": "Array of message entries",
      "default": null,
      "items": {
        "type": "object",
        "properties": {
          "text": {
            "type": "string",
            "description": "Text part of the user message",
            "default": null
          }
        }
      }
    },
    "prompt": {
      "type": "string",
      "description": "ID of the prompt to generate an additional text based user message",
      "examples": ["search-prompt"],
      "default": null
    },
    "input": {
      "type": "object",
      "description": "Data input elements to be used for populating prompt",
      "examples": [{"search": "keyword"}],
      "default": null
    }
  }
}

Base64 contents can also be passed as data URI values with the correct prefix.

With event metadata parameters as:

Parameters (table)

Parameter	Definition	Example	Default
Message Pattern	Used for tool sagas, allowing merging data from original event to saga call, with arguments as agent-generated input and payload as original	merge(arguments, {user: payload.user})	-
Full Result	Whether response should include full AI call details, such as token counts and tool executions	true	false
Json Response	Whether model response should be automatically parsed as a JSON object	true	false

Parameters (JSON Schema)

LLMGenerateImage

Performs an image generation with a target LLM model provider. Event metadata fields applicable for this action are as follows:

Fields (table)

Field	Definition	Example	Default
Domain	ID of the model to use	dalle_gen	-
Input Element	Json path for the input in request event payload	{prompt: "", imageCount: ""}	-
Output Element	Json path for the output in response event payload	imageBase64	-

Fields (JSON Schema)

With event metadata parameters as:

Parameters (table)

Parameter	Definition	Example	Default
Full Result	Whether response should include full AI call details	true	false

Parameters (JSON Schema)

LLMEditImage

Performs an image edit with a target LLM model provider. Event metadata fields applicable for this action are as follows:

Fields (table)

Field	Definition	Example	Default
Domain	ID of the model to use	dalle_gen	-
Input Element	Json path for the input in request event payload	{prompt: "", image: {revisedPrompt: "", base64Data: "", mimeType: "", url: ""}, mask: ""}	-
Output Element	Json path for the output in response event payload	output	-

Fields (JSON Schema)

With event metadata parameters as:

Parameters (table)

Parameter	Definition	Example	Default
Full Result	Whether response should include full AI call details	true	false

Parameters (JSON Schema)