> For clean Markdown of any page, append .md to the page URL. > For a complete documentation index, see https://docs.nvidia.com/local-ai/nvpair/llms.txt. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/local-ai/nvpair/_mcp/server. # OpenAI-compatible chat completion POST http%3A%2F%2F127.0.0.1%3A11434/v1/chat/completions Content-Type: application/json Send an OpenAI-compatible chat completion request to the local PAIR proxy. PAIR forwards the entire request to a single eligible node and returns the engine's response. Reference: https://docs.nvidia.com/local-ai/nvpair/pair-api/open-ai-compatible/create-chat-completion ## Request ### Body (application/json) This endpoint expects a ChatCompletionRequest. - `model` (string, required) — Model name that must be available on an eligible node. - `messages` (list of ChatMessage, required) - `stream` (boolean, optional, default: false) - `temperature` (float, optional) - `max_tokens` (integer, optional) ## Response ### 200 Chat completion from the serving node. - `id` (string, required) - `object` (string, required) - `created` (long, required) — Unix timestamp in seconds. - `model` (string, required) - `choices` (list of ChatCompletionChoice, required) - `usage` (Usage, optional) ## Types ### ChatMessage - `role` (enum, required) - Allowed values: `system`, `user`, `assistant`, `tool` - `content` (string, required) ### ChatCompletionChoice - `index` (integer, required) - `message` (ChatMessage, required) - `finish_reason` (enum, optional) - Allowed values: `stop`, `length`, `tool_calls`, `content_filter` ### Usage - `prompt_tokens` (integer, optional) - `completion_tokens` (integer, optional) - `total_tokens` (integer, optional) ## Examples **Request** ```json { "model": "llama3.1", "messages": [ { "role": "user", "content": "Explain what NVIDIA Personal AI Router does in one sentence." } ] } ``` **Response** ```json { "id": "chatcmpl-01J0D5P0KZESQHTR2HWK7S5W27", "object": "chat.completion", "created": 1780000000, "model": "llama3.1", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "NVIDIA Personal AI Router routes each AI request to one eligible system on your local network to serve inference." }, "finish_reason": "stop" } ], "usage": { "prompt_tokens": 18, "completion_tokens": 24, "total_tokens": 42 } } ``` **SDK Code** ```python default import requests url = "https://api.example.com/v1/chat/completions" payload = { "model": "llama3.1", "messages": [ { "role": "user", "content": "Explain what NVIDIA Personal AI Router does in one sentence." } ] } headers = {"Content-Type": "application/json"} response = requests.post(url, json=payload, headers=headers) print(response.json()) ``` ```javascript default const url = 'https://api.example.com/v1/chat/completions'; const options = { method: 'POST', headers: {'Content-Type': 'application/json'}, body: '{"model":"llama3.1","messages":[{"role":"user","content":"Explain what NVIDIA Personal AI Router does in one sentence."}]}' }; try { const response = await fetch(url, options); const data = await response.json(); console.log(data); } catch (error) { console.error(error); } ``` ```go default package main import ( "fmt" "strings" "net/http" "io" ) func main() { url := "https://api.example.com/v1/chat/completions" payload := strings.NewReader("{\n \"model\": \"llama3.1\",\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Explain what NVIDIA Personal AI Router does in one sentence.\"\n }\n ]\n}") req, _ := http.NewRequest("POST", url, payload) req.Header.Add("Content-Type", "application/json") res, _ := http.DefaultClient.Do(req) defer res.Body.Close() body, _ := io.ReadAll(res.Body) fmt.Println(res) fmt.Println(string(body)) } ``` ```ruby default require 'uri' require 'net/http' url = URI("https://api.example.com/v1/chat/completions") http = Net::HTTP.new(url.host, url.port) http.use_ssl = true request = Net::HTTP::Post.new(url) request["Content-Type"] = 'application/json' request.body = "{\n \"model\": \"llama3.1\",\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Explain what NVIDIA Personal AI Router does in one sentence.\"\n }\n ]\n}" response = http.request(request) puts response.read_body ``` ```java default import com.mashape.unirest.http.HttpResponse; import com.mashape.unirest.http.Unirest; HttpResponse response = Unirest.post("https://api.example.com/v1/chat/completions") .header("Content-Type", "application/json") .body("{\n \"model\": \"llama3.1\",\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Explain what NVIDIA Personal AI Router does in one sentence.\"\n }\n ]\n}") .asString(); ``` ```php default request('POST', 'https://api.example.com/v1/chat/completions', [ 'body' => '{ "model": "llama3.1", "messages": [ { "role": "user", "content": "Explain what NVIDIA Personal AI Router does in one sentence." } ] }', 'headers' => [ 'Content-Type' => 'application/json', ], ]); echo $response->getBody(); ``` ```csharp default using RestSharp; var client = new RestClient("https://api.example.com/v1/chat/completions"); var request = new RestRequest(Method.POST); request.AddHeader("Content-Type", "application/json"); request.AddParameter("application/json", "{\n \"model\": \"llama3.1\",\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Explain what NVIDIA Personal AI Router does in one sentence.\"\n }\n ]\n}", ParameterType.RequestBody); IRestResponse response = client.Execute(request); ``` ```swift default import Foundation let headers = ["Content-Type": "application/json"] let parameters = [ "model": "llama3.1", "messages": [ [ "role": "user", "content": "Explain what NVIDIA Personal AI Router does in one sentence." ] ] ] as [String : Any] let postData = JSONSerialization.data(withJSONObject: parameters, options: []) let request = NSMutableURLRequest(url: NSURL(string: "https://api.example.com/v1/chat/completions")! as URL, cachePolicy: .useProtocolCachePolicy, timeoutInterval: 10.0) request.httpMethod = "POST" request.allHTTPHeaderFields = headers request.httpBody = postData as Data let session = URLSession.shared let dataTask = session.dataTask(with: request as URLRequest, completionHandler: { (data, response, error) -> Void in if (error != nil) { print(error as Any) } else { let httpResponse = response as? HTTPURLResponse print(httpResponse) } }) dataTask.resume() ```