Actions18
- ChatGPT Actions
- Claude Actions
- Gemini Actions
- Keyword Data Actions
- LLM Mentions Actions
- Perplexity Actions
ChatGPT → Get Live ChatGPT LLM Responses
AI-generatedSummary
Send a prompt and optional conversation history to a selected ChatGPT model through DataForSEO’s live LLM responses endpoint. Optional settings control response length, sampling, and web search.
Inputs
- Prompt for the AI Model (required) — Question or task for the model; the UI hint allows up to 500 characters.
- Name of the AI Model (required) — Model identifier. A basic model name may resolve to its latest version; the UI points to DataForSEO’s models endpoint for available models.
- Maximum Number of Tokens in the AI Response — Optional response token limit.
- Randomness of the AI Response — Optional temperature from 0 to 2.
- Diversity of the AI Response — Optional top_p value from 0 to 2.
- Enable Web Search? — Enables web search; defaults to false.
- Force AI Agent to Use Web Search? — Sent only when web search is enabled, to request that the model use it.
- ISO Country Code of the Location — Optional country code for web search. The UI says this is required when a city is specified.
- City Name of the Location — Optional location for web search.
- Instructions for the AI Behaviour — Optional system message defining the model’s role, tone, or behavior.
- Conversation History — Optional sequence of messages, each with a required role (User or AI) and message text. The UI marks the collection required, but it defaults to empty and the implementation sends null when it has no messages.
Output shape
A single DataForSEO API response object per input item, wrapped as an n8n item’s json value.
The node sends one request to
/ai_optimization/chat_gpt/llm_responses/livefor each input item and returns the API response as received; the source does not reshape it into just the generated text. Optional numeric fields and blank strings are sent as null. The force-web-search setting is sent only when web search is enabled.