Chat Completions

Gateway provides OpenAI compatible Chat Completions API endpoints, letting you use multiple AI providers through a familiar interface. You can use existing OpenAI client libraries, switch to Gateway with a URL change, and keep your current tools and workflows without code rewrites.

The Chat Completions API implements the same specification as the OpenAI Chat Completions API.

Endpoint: POST https://api.thesys.dev/v1/embed/chat/completions

Configure the client

import OpenAI from "openai";

export const gateway = new OpenAI({
  apiKey: process.env.THESYS_API_KEY,
  baseURL: "https://api.thesys.dev/v1/embed",
});

See Authentication before using the client in a server route.

Generate OpenUI Lang

Configure the system message for OpenUI generation, then send the same messages shape used by the OpenAI SDK:

import { generateSystemPrompt } from "@openuidev/lang-core";
import { gateway } from "./lib/gateway";
import library from "./openui.spec.json";

const completion = await gateway.chat.completions.create({
  model: "openai/gpt-5",
  messages: [
    { role: "system", content: generateSystemPrompt({ cloud: true, library }) },
    { role: "user", content: "Compare quarterly revenue by region." },
  ],
  stream: true,
});

The assistant message contains OpenUI Lang. When the request is configured for OpenUI generation, the Gateway validates and corrects that language while it streams.

Plain text requests

The endpoint also supports ordinary model traffic. Send your own system prompt without OpenUI generation instructions, and the Gateway forwards a text-oriented request without applying OpenUI Lang correction.

const completion = await gateway.chat.completions.create({
  model: "anthropic/claude-sonnet-5",
  messages: [{ role: "user", content: "Summarize this support ticket." }],
});

Unknown providers can route through OpenRouter where supported. Use Models for the standard provider naming convention.

Stream to an agent UI

Relay the server response without changing its SSE shape. In the browser, configure Agent Interface adapters and message formats for the Chat Completions stream.

Chat Completions is message-based. Your application owns conversation history and sends the relevant system, user, assistant, and tool messages on every turn. Use the Responses API when you want Gateway-managed persistent conversations.

Function tools

Chat Completions returns function tool calls to the application; it does not execute them. Run the standard loop:

1. Send messages and function declarations.
2. Read tool_calls from the assistant message.
3. Execute each function in your application.
4. Append the assistant tool-call message and each tool result.
5. Continue until the model returns the final OpenUI Lang response.