> ## Documentation Index > Fetch the complete documentation index at: https://aisa.one/docs/llms.txt > Use this file to discover all available pages before exploring further. # Live Perplexity LLM Responses > Live Perplexity LLM Responses endpoint allows you to retrieve structured responses from a specific Perplexity AI model, based on the input parameters. ## OpenAPI ````yaml openapi/dataforseo.json POST /dataforseo/ai_optimization/perplexity/llm_responses/live openapi: 3.0.3 info: title: DataForSEO API version: 1.0.0 description: DataForSEO API endpoints exposed through the AIsa unified gateway. servers: - url: https://api.aisa.one/apis/v1 security: - BearerAuth: [] paths: /dataforseo/ai_optimization/perplexity/llm_responses/live: post: summary: Live Perplexity LLM Responses description: >- Live Perplexity LLM Responses endpoint allows you to retrieve structured responses from a specific Perplexity AI model, based on the input parameters. operationId: post_dataforseo_ai_optimization_perplexity_llm_responses_live requestBody: required: true content: application/json: schema: type: object properties: user_prompt: type: string description: >- prompt for the AI model required field the question or task you want to send to the AI model; you can specify up to 500 characters in the user_prompt field model_name: type: string description: >- name of the AI model required field model_nameconsists of the actual model name and version name; if the basic model name is specified, its latest version will be set by default; you can receive the list of available LLM models by making a separate request to the following endpoint: https://api.dataforseo.com/v3/ai_optimization/perplexity/llm_responses/models max_output_tokens: type: integer description: >- maximum number of tokens in the AI response optional field minimum value: 1 maximum value: 4096; default value: 2048; Note: if the reasoning model is specified in the request, the output token count may exceed the specified max_output_tokens limit temperature: type: number description: >- randomness of the AI response optional field higher values make output more diverse lower values make output more focused minimum value: 0 maximum value: 1.9 default value: 0.77 top_p: type: number description: >- diversity of the AI response optional field controls diversity of the response by limiting token selection minimum value: 0 maximum value: 1 default value: 0.9 web_search_country_iso_code: type: string description: >- country code for web search localization optional field specify the country ISO code to get localized web search results Note: available only for Perplexity Sonar models example: US system_message: type: string description: >- instructions for the AI behavior optional field defines the AI's role, tone, or specific behavior you can specify up to 500 characters in the system_message field message_chain: type: array items: type: string description: >- conversation history optional field array of message objects representing previous conversation turns; each object must contain: role string with either user or ai role; message string with message content (max 500 characters); you can specify maximum of 10 message objects in the array; Note: for Perplexity models, messages must strictly alternate between user and AI roles (user → ai); example: "message_chain": [{"role":"user","message":"Hello, what’s up?"},{"role":"ai","message":"Hello! I’m doing well, thank you. How can I assist you today?"}] tag: type: string description: >- user-defined task identifier optional field the character limit is 255 you can use this parameter to identify the task and match it with the result you will find the specified tag value in the data object of the response required: - user_prompt - model_name responses: '200': description: Successful response content: application/json: schema: type: object properties: version: type: string description: the current version of the API status_code: type: integer description: >- general status code you can find the full list of the response codes here Note: we strongly recommend designing a necessary system for handling related exceptional or error conditions status_message: type: string description: >- general informational message you can find the full list of general informational messages here time: type: string description: execution time, seconds cost: type: number description: total tasks cost, USD tasks_count: type: integer description: the number of tasks in the tasks array tasks_error: type: integer description: >- the number of tasks in the tasks array returned with an error tasks: type: array items: type: string description: array of tasks tasks.id: type: string description: >- task identifier unique task identifier in our system in the UUID format tasks.status_code: type: integer description: >- status code of the task generated by DataForSEO; can be within the following range: 10000-60000 you can find the full list of the response codes here tasks.status_message: type: string description: >- informational message of the task you can find the full list of general informational messages here tasks.time: type: string description: execution time, seconds tasks.cost: type: number description: >- cost of the task, USD includes the base task price plus the money_spent value tasks.result_count: type: integer description: number of elements in the result array tasks.path: type: array items: type: string description: URL path tasks.data: type: object description: >- contains the same parameters that you specified in the POST request tasks.result: type: array items: type: string description: array of results tasks.result.model_name: type: string description: name of the AI model used tasks.result.input_tokens: type: integer description: >- number of tokens in the input total count of tokens processed tasks.result.output_tokens: type: integer description: >- number of tokens in the output total count of tokens generated in the AI response tasks.result.web_search: type: boolean description: >- indicates if web search was used Note: web search is enabled by default in Perplexity Sonar models tasks.result.money_spent: type: number description: >- cost of AI tokens, USD the price charged by the third-party AI model provider for according to its Pricing tasks.result.datetime: type: string description: >- date and time when the result was received in the UTC format: “yyyy-mm-dd hh-mm-ss +00:00” example: 2019-11-15 12:57:46 +00:00 tasks.result.items: type: array items: type: string description: >- array of response items contains structured AI response data tasks.result.items.type: type: string description: type of the element = 'message' tasks.result.items.sections: type: array items: type: string description: >- array of content sections contains different parts of the AI response tasks.result.items.sections.type: type: string description: type of element='text' tasks.result.items.sections.text: type: string description: AI-generated text content tasks.result.items.sections.annotations: type: array items: type: string description: >- array of references used to generate the response equals null if the web_search parameter is not set to true Note: annotations may return empty even when web_search is true, as the AI will attempt to retrieve web information but may not find relevant results tasks.result.items.sections.annotations.title: type: string description: the domain name or title of the quoted source tasks.result.items.sections.annotations.url: type: string description: URL of the quoted source tasks.result.fan_out_queries: type: array items: type: string description: >- array of fan-out queries contains related search queries derived from the main query to provide a more comprehensive response '400': description: Bad request '401': description: Unauthorized '429': description: Rate limit exceeded '500': description: Internal server error components: securitySchemes: BearerAuth: type: http scheme: bearer ````