@assistant-hub/sdk

Client SDK for Assistant Hub AI Gateway.

Drop-in replacement for the OpenAI SDK — just change the base URL and API key.

Install

npm install @assistant-hub/sdk

Environment Variables

The SDK reads these automatically — no configuration needed:

ASSISTANT_HUB_API_KEY=sk_hub_...
ASSISTANT_HUB_BASE_URL=https://api.aichat.new   # optional, this is the default

Quick Start

import { HubClient } from "@assistant-hub/sdk";

// Reads ASSISTANT_HUB_API_KEY and ASSISTANT_HUB_BASE_URL from env
const hub = new HubClient();

const response = await hub.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello!" }],
});

console.log(response.choices[0].message.content);

Vercel AI SDK Integration

Use @assistant-hub/sdk/ai-sdk for seamless integration with the Vercel AI SDK:

import { createHub } from "@assistant-hub/sdk/ai-sdk";
import { generateText } from "ai";

// Reads env vars automatically
const hub = createHub();

const { text } = await generateText({
  model: hub("gpt-4o"),
  prompt: "Hello!",
});

Or use the pre-configured default instance:

import { hub } from "@assistant-hub/sdk/ai-sdk";
import { streamText } from "ai";

const result = streamText({
  model: hub("claude-sonnet-4-20250514"),
  prompt: "Write a haiku",
});

You can also pass options explicitly:

const hub = createHub({
  apiKey: "sk_hub_...",
  baseURL: "https://your-gateway.example.com/v1",
});

Streaming

const stream = await hub.chat.completions.create({
  model: "claude-sonnet-4-20250514",
  messages: [{ role: "user", content: "Write a haiku" }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}

Embeddings

const embeddings = await hub.embeddings.create({
  model: "text-embedding-3-small",
  input: "Hello world",
});

List Models

const models = await hub.models.list();

Context Builders

Chain context for request tracking:

const response = await hub
  .withUser("user-123")
  .withMetadata({ session: "abc" })
  .withProvider("openai")
  .chat.completions.create({
    model: "gpt-4o",
    messages: [{ role: "user", content: "Hello" }],
  });

Feedback & Scores

Submit a single feedback score:

await hub.feedback("hrl_xxx", 5);

Submit multiple scores at once (including boolean scores):

await hub.scores("hrl_xxx", {
  quality: 5,
  relevance: 3,
  thumbs_up: true,
});

Configuration

const hub = new HubClient({
  apiKey: "sk_hub_...",                           // or ASSISTANT_HUB_API_KEY env var
  baseURL: "https://your-gateway.aichat.new",     // or ASSISTANT_HUB_BASE_URL env var
  timeout: 30_000,                                // optional, default 60s
  defaultHeaders: { "X-Custom": "value" },        // optional
});

Works with OpenAI SDK

You can also use the official OpenAI SDK directly:

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ASSISTANT_HUB_API_KEY,
  baseURL: "https://api.aichat.new/v1",
});

License

MIT