GraphQL AI Text Mutations

The AI Text mutations provide AI-powered content generation, refinement, translation, transcription, and file description capabilities through the GraphQL API. All mutations require authentication.

Generate text content from a prompt, optionally using existing files as reference context.

mutation {
  write(prompt: "Write a product description for a wireless headphone", context: "tech blog, casual tone", files: ["file-id-1"])
}

Arguments:

Argument	Type	Description
`prompt`	`String!`	Text description of the content to generate (required)
`context`	`String`	Additional context to guide the writing style and tone
`files`	`[String!]`	List of file IDs to use as reference material

Returns: String — The generated text content.

Example response:

{
  "data": {
    "write": "Experience crystal-clear audio with our latest wireless headphones..."
  }
}

Refine existing page content using AI. Pass a prompt describing the desired changes along with the current content structure.

mutation {
  refine(
    prompt: "Make the tone more professional and add a call to action",
    content: [{"type": "text", "data": {"text": "Check out our cool stuff!"}}],
    type: "text",
    context: "corporate landing page"
  )
}

Arguments:

Argument	Type	Description
`prompt`	`String!`	Text description of the desired refinements (required)
`content`	`JSON!`	The existing content elements to refine (required)
`type`	`String`	The content element type to focus on
`context`	`String`	Additional context to guide the refinement

Returns: JSON — The refined content elements in the same structure as the input.

Example response:

{
  "data": {
    "refine": [
      {
        "type": "text",
        "data": {
          "text": "Discover our premium solutions designed to elevate your business. Contact us today to schedule a consultation."
        }
      }
    ]
  }
}

Generate complete pages and content structures from prompts, optionally using existing files as reference.

mutation {
  synthesize(
    prompt: "Create a landing page for a SaaS product with hero, features, and pricing sections",
    context: "modern tech startup",
    files: ["file-id-1", "file-id-2"]
  )
}

Arguments:

Argument	Type	Description
`prompt`	`String!`	Text description of the page or content to generate (required)
`context`	`String`	Additional context to guide the generation
`files`	`[String!]`	List of file IDs to use as reference material

Returns: String — The generated page content.

Example response:

{
  "data": {
    "synthesize": "[{\"type\":\"heading\",\"data\":{\"text\":\"Streamline Your Workflow\"}},...]"
  }
}

Generate a text summary or description of a file's content using AI. Useful for auto-generating alt text for images or descriptions for documents.

mutation {
  describe(file: "0187d6ab-b76d-75ee-8b0d-1b59cc3a1ab7", lang: "en")
}

Arguments:

Argument	Type	Description
`file`	`String!`	The ID of the file to describe (required)
`lang`	`String`	ISO language code for the description output

Returns: String — A text description of the file content.

Example response:

{
  "data": {
    "describe": "A panoramic photograph of a coastal city at dusk, featuring illuminated buildings along the waterfront with mountains in the background."
  }
}

Transcribe an uploaded audio file to text using AI speech recognition.

mutation Transcribe($file: Upload!) {
  transcribe(file: $file)
}

Arguments:

Argument	Type	Description
`file`	`Upload!`	The audio file to transcribe (required)

Returns: JSON — The transcription result.

Example response:

{
  "data": {
    "transcribe": {
      "en": "Welcome to our podcast. Today we are discussing the future of content management systems..."
    }
  }
}

Translate one or more text strings to a target language using AI.

mutation {
  translate(
    texts: ["Welcome to our website", "Learn more about our products"],
    to: "de",
    from: "en",
    context: "corporate website"
  )
}

Arguments:

Argument	Type	Description
`texts`	`[String!]!`	Array of text strings to translate (required)
`to`	`String!`	Target ISO language code, e.g. `de`, `fr`, `es` (required)
`from`	`String`	Source ISO language code (auto-detected if omitted)
`context`	`String`	Additional context to improve translation quality

Returns: [String!]! — Array of translated strings in the same order as the input.

Example response:

{
  "data": {
    "translate": [
      "Willkommen auf unserer Website",
      "Erfahren Sie mehr über unsere Produkte"
    ]
  }
}