Generate text variations using AI

post

https://api.convert.com/api/v2/accounts/{account_id}/ai/text

Leverages Generative AI to create alternative text versions (variants) for a given piece of content. Users can specify the number of variations, the original content, an AI model framework (e.g., Cialdini, AIDA), and a specific principle within that framework. This helps in quickly generating creative options for A/B testing headlines, calls to action, or other textual elements. Optionally, an OpenAI Key can be provided for using more advanced models or larger context.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Path Params

account_id

integer

required

ID of the account under which the api call executes

Body Params

Base structure for requesting AI-generated text variations. Specifies the content to rewrite, the AI model framework and principle to use, number of variations, and optional contextual content.

parameters

object

required

Parameters to customize the AI generation process.

content

string

required

The original text content (e.g., headline, call-to-action, product description) that you want the AI to rewrite or generate variations for.

openAI_Key

string

(Optional) Your OpenAI API key. Providing this may allow for using more advanced OpenAI models (e.g., GPT-4) or handling larger context_content, potentially leading to higher quality or more nuanced variations. If not provided, Convert's default AI capabilities will be used.

context_content

string

(Optional) Additional text to provide context to the AI model, helping it understand the subject matter, tone, or specific constraints for the generated variations. For example, you could provide the product page content if generating a headline for that product. If no openAI_Key is provided, this content may be trimmed to the first 200 words. When principle_class=none is provided, this field can be used for custom prompt instructions.

previousResponses

array of strings

(Optional) A list of text strings representing previously generated AI variants for the same input content. Providing these helps the AI avoid generating duplicate or very similar suggestions, ensuring more diverse outputs if requesting variations iteratively.

previousResponses

principle_class

enum

Framework model from which principles used for text generation are selected

Allowed:

Responses

200A list of AI-generated text variations based on the provided input content, model framework, and principle. Each variant is a string.

defaultIndicates an error occurred while processing the request. The code provides an HTTP status code, message offers a human-readable explanation or an array of validation errors, and fields (if present) specifies which input fields were problematic.