Streaming

Streaming enables NovaAI to return generated content in real time instead of waiting for the entire response to complete. This improves responsiveness and creates a smoother user experience for chat applications, AI assistants, and interactive interfaces.

Overview

With streaming enabled, the model sends small chunks of generated text as they become available. Your application can display these chunks immediately, reducing perceived latency and making conversations feel more natural.

Note: Streaming is supported by compatible NovaAI chat and agent models. Check the model documentation before enabling this feature.

How Streaming Works

Rendering diagram…

Enable Streaming

1const stream = await client.chat.create({
2  model: "nova-chat-4",
3  stream: true,
4  messages: [
5    {
6      role: "user",
7      content: "Write a short introduction to artificial intelligence."
8    }
9  ]
10});
11 
12for await (const chunk of stream) {
13  process.stdout.write(chunk.delta);
14}

Streaming Properties

stream

boolean

required

Enables real-time response streaming.

model

string

required

Model used for generation.

messages

array

required

Conversation history sent to the model.

Common Use Cases

AI chat interfaces
Writing assistants
Live code generation
Customer support bots
Real-time content generation

Tip: Streaming provides a better user experience for long responses by displaying content as it's generated rather than waiting for the complete output.

Next Step

Continue to Function Calling to learn how NovaAI models can invoke external tools and APIs to perform actions beyond text generation.

Streaming

Overview

How Streaming Works

Enable Streaming

Streaming Properties

Common Use Cases

Next Step

Related Documentaion

Chat Completion

Audio

Function Calling