
One call sends a prompt to a specified LLM model and returns the model's text response. Each item in the answer is a message section containing the model's reply text. The answer contained 1 message with 1 text section.
1 live route on DataForSEO, from $0.00071 per call. Every route below answered our last live test on 2026-10-06; the gateway re-tests them daily.
The upstream price you are metered at, cheapest first. A route without a published price says so.
| Provider | Endpoint | Price |
|---|---|---|
ChatGPT API answer to a promptPOST ai_optimization/chat_gpt/llm_responses/live | $0.00071$0.71 / 1k calls |
From a real call to the route: DataForSEO.
result[].items[].typeType of response (e.g., 'message')result[].items[].sections[].textThe model's text response to the promptresult[].items[].sections[].typeSection type (e.g., 'text')result[].model_nameThe LLM model used (e.g., 'gpt-4o-mini-2024-07-18')result[].input_tokensNumber of tokens in the input promptresult[].output_tokensNumber of tokens in the model's responseresult[].datetimeTimestamp of the responseHTTP with your key in x-api-key, or the call_endpoint tool of our MCP server.
| Parameter | Type | Notes | |
|---|---|---|---|
| model_name | str · body | JSON body field | e.g. "gpt-4o-mini" |
| user_prompt | str · body | JSON body field | e.g. "Where can I buy Pokemon cards online?" |
curl -X POST "https://fetchapi.co/api/gw/dataforseo/ai_optimization/chat_gpt/llm_responses/live" \
-H "x-api-key: YOUR_KEY" -H "content-type: application/json" \
-d '[{"model_name": "gpt-4o-mini", "user_prompt": "Where can I buy Pokemon cards online?"}]'call_endpoint("dataforseo.ai_optimization_chat_gpt_llm_responses", {}, [{"model_name": "gpt-4o-mini", "user_prompt": "Where can I buy Pokemon cards online?"}]){
"cost": 0.00071,
"time": "2.6819 sec.",
"tasks": [
{
"id": "10040440-2697-0612-0000-876ee1e642bd",
"cost": 0.00071,
"data": {
"se": "chat_gpt",
"api": "ai_optimization",
"function": "llm_responses",
"model_name": "gpt-4o-mini",
"user_prompt": "Where can I buy Pokemon cards online?"
},
"path": [
"v3",
"ai_optimization",
"… 2 more"
],
"time": "2.6759 sec.",
"result": [
{
"items": [
{
"type": "…",
"sections": "…"
}
],
"datetime": "2026-10-04 04:40:11 +00:00",
"model_name": "gpt-4o-mini-2024-07-18",
"web_search": false,
"money_spent": 0.0001095,
"input_tokens": 26,
"output_tokens": 176,
"fan_out_queries": null,
"reasoning_tokens": 0
}
],
"status_code": 20000,
"result_count": 1,
"status_message": "Ok."
}
],
"version": "0.1.20260917",
"status_code": 20000,
"tasks_count": 1,
"tasks_error": 0,
"status_message": "Ok."
}Call it over HTTP or from any MCP client. Each key has a rate limit and a monthly cap; calls are metered at the upstream price.