POST
/
generate
/
text
curl --location 'https://api.mixpeek.com/generate/text' \
--header 'Authorization: Bearer API_KEY' \
--header 'Content-Type: application/json' \
--data '{
    "model": {
        "provider": "GPT",
        "model": "gpt-4-turbo-preview"
    },
    "response_format": {
        "type": "object",
        "properties": {
            "city": {
                "title": "City",
                "type": "string"
            },
            "neighborhood": {
                "title": "Neighborhood",
                "type": "string"
            },
            "weather": {
                "title": "Weather",
                "type": "number"
            }
        },
        "required": ["city", "neighborhood", "weather"]
    },
    "context": "Please tell me the weather and make sure to respond in the provided JSON schema",
    "messages": [],
    "settings": {}
}'

{
  "response": {},
  "metadata": {},
  "elapsed_time": 123
}

​
Request

model
object
required

Specifies the language model provider and the specific model to be used for generation. This field allows customization of the text generation process based on the model’s capabilities and characteristics.

response_format
object
required

Defines the JSON schema for the response. This schema specifies the structure and types of the data that the generated text should conform to, ensuring that the output meets specific data format requirements.

context
string
required

The context or prompt based on which the model will generate the text. This should be a clear and concise statement or question that guides the model’s output within the constraints of the specified response format.

messages
array

An optional array of messages or additional context that can be provided to the model to further influence or guide the generation process.

settings
object

Additional settings for the generation process, allowing for customization of model behavior, such as adjusting creativity levels, specifying output length, or other model-specific parameters.

​
Response

response
object
required

The generated text output structured according to the specified JSON schema. This object contains the data fields and values as defined in the response format.

metadata
object
required

Metadata about the generation process, including the total number of tokens generated, the unique identifier for the generation task, and details about the model used.

elapsed_time
number
required

The time taken to generate the response, measured in milliseconds.

ExtractEmbed