Skip to main content
Generate a model response for a given conversation. This endpoint lets you build interactive chat applications using AI models like openai/gpt-4, openai/gpt-4o, or even neuroa/m1-preview. It mirrors the behavior of OpenAI’s chat completions API. 
const { Neuroa } = require('neuroa-sdk');

const client = new Neuroa({
  apiKey: process.env.NEUROA_API_KEY
});

// Basic completion (non-streaming)
client.models.generate.completions({
  model: "openai/gpt-4.1",
  messages: [
    { role: "user", content: "Hello, world!" }
  ]
}).then(response => {
  console.log(response.choices[0].message.content);
}).catch(error => {
  console.error("Completions generation error:", error);
});

​
Request Body

Send a list of messages and specify the model you’d like to use. 
{
  "model": "openai/gpt-4.1",
  "messages": [
    { "role": "user", "content": "Tell me a fun fact about space." }
  ]
}
FieldTypeRequiredDescription
modelstring✅ YesModel ID to use (e.g. openai/o1-pro, anthropic/claude-4-opus)
messagesarray✅ YesOrdered history of messages in the chat
temperaturenumber❌ NoControls randomness: 0 (deterministic) to 2 (more random)
streamboolean❌ NoEnables streaming if true
max_tokensnumber❌ NoMax number of tokens to generate
thinkingThinkingObject❌ NoOnly supported by M1 Series models
toolsobject❌ NoTool calling payload (OpenAI, Gemini, Anthropic support this)

​
thinking Object

Optional parameter for M1 Series models. 
"thinking": {
  "thinking": true,
  "effort": "high"
}
FieldTypeRequiredDescription
thinkingboolean✅ YesEnables internal thinking-like behavior
effortstring✅ YesCan be "low", "medium" or "high"

​
Example Response

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1677858242,
  "model": "openai/gpt-4.1",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "A fun fact: One day on Venus is longer than one year on Venus!"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 14,
    "completion_tokens": 17,
    "total_tokens": 31
  }
}

​
Streaming Responses

When stream: true is set, the API responds with a Server-Sent Events (SSE) stream. Each chunk includes a delta until completion. Sample stream response: 
data: {
  "choices": [
    {
      "delta": {
        "content": "Venus"
      }
    }
  ]
}

data: {
  "choices": [
    {
      "delta": {
        "content": " is"
      }
    }
  ]
}
...
data: [DONE]
Each data: message is a part of the response. The client should collect delta.content pieces to reconstruct the final message.

​
Authentication

Use the Bearer token method in your Authorization header: 
Authorization: Bearer YOUR_API_KEY

​
Need More?

Want to build full real-time chat apps? Check out the OpenAI streaming guide or Neuroa’s SDK documentation for richer control.