Chat API

The Chat API is the primary interface for working with Synoria models, allowing for text generation, question answering, and content summarization. Each request specifies a model and includes messages to effectively guide the conversation. You can adjust parameters like temperature and output tokens to influence creativity and response length. This comprehensive guide covers request structure, customization options, and response management.

Create a Chat Request

A chat request consists of a model and a list of messages. Each message includes a role and content, allowing the model to understand the conversation and generate relevant responses.

import client from "./client.js";

async function createChat() {
  const response = await client.chat.generate({
    model: "axiom-1",
    messages: [
      {
        role: "user",
        content: "Explain how neural networks work."
      }
    ]
  });

  console.log(response.output);
}

createChat();

Customize Model Behavior

Fine-tune responses by adjusting generation parameters. These settings help balance creativity, consistency, response length, and output quality for different use cases.

import client from "./client.js";

const response = await client.chat.generate({
  model: "axiom-1",
  temperature: 0.8,
  maxTokens: 500,
  topP: 0.95,
  frequencyPenalty: 0.2,
  presencePenalty: 0.1,
  messages: [
    {
      role: "system",
      content: "You are an expert software architect."
    },
    {
      role: "user",
      content: "Design a scalable notification service."
    }
  ]
});

console.log(response.output);

Handle the Response

Every response contains the generated output along with useful metadata such as the model and token usage. These values help monitor performance and optimize costs.

const result = await client.chat.generate({
  model: "axiom-1",
  messages: [
    {
      role: "user",
      content: "Write a product tagline."
    }
  ]
});

console.log(result.output);
console.log(result.model);
console.log(result.usage.total_tokens);

Continue Exploring

The Chat API provides a flexible foundation for building AI-powered assistants, productivity tools, coding experiences, and customer support applications. Next, you'll learn how to choose the right Synoria model for different workloads and performance requirements.