SDKs & tools

Everything that speaks the OpenAI format works with a base URL and a key. We have verified these configurations.

Official OpenAI SDKs

Install the SDK for your language and set the base URL — the rest is standard.

Python

pip install openai

JavaScript / TypeScript

npm install openai

Base URL for every tool: https://api.futureway.ai/v1. The key comes from the FUTUREWAY_API_KEY environment variable.

OpenCode

OpenCode is a coding agent in the terminal. With this configuration it uses our models — your code never leaves the EU.

1. Install (macOS, Linux, WSL)

curl -fsSL https://opencode.ai/install | bash

2. opencode.json in the project or under ~/.config/opencode/

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "futureway": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "FutureWay AI",
      "options": {
        "baseURL": "https://api.futureway.ai/v1",
        "apiKey": "{env:FUTUREWAY_API_KEY}"
      },
      "models": {
      "futureway-fast": { "name": "Mistral Small 3.2 24B Instruct" },
      "futureway-smart": { "name": "Mistral Medium 3.5 128B" },
      "futureway-bigthink": { "name": "Qwen3.5 397B A17B" }
      }
    }
  }
}

3. Run

opencode run --model futureway/futureway-smart "Erkläre mir diese Codebasis in fünf Sätzen."

Continue

Continue (VS Code, JetBrains) — chat, edit and embeddings through our API, key from Continue's secrets:

name: FutureWay AI
version: 0.0.1
models:
  - name: FutureWay futureway-smart
    provider: openai
    model: futureway-smart
    apiBase: https://api.futureway.ai/v1
    apiKey: ${{ secrets.FUTUREWAY_API_KEY }}
    roles: [chat, edit, apply]
  - name: FutureWay Embeddings
    provider: openai
    model: futureway-embed
    apiBase: https://api.futureway.ai/v1
    apiKey: ${{ secrets.FUTUREWAY_API_KEY }}
    roles: [embed]

LangChain

LangChain uses the OpenAI integration with a changed base URL:

import os
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="futureway-smart",
    base_url="https://api.futureway.ai/v1",
    api_key=os.environ["FUTUREWAY_API_KEY"],
)
print(llm.invoke("Fasse den EU AI Act in drei Sätzen zusammen.").content)

Vercel AI SDK

The Vercel AI SDK through the OpenAI-compatible provider:

import { createOpenAICompatible } from "@ai-sdk/openai-compatible";
import { generateText } from "ai";

const futureway = createOpenAICompatible({
  name: "futureway",
  baseURL: "https://api.futureway.ai/v1",
  apiKey: process.env.FUTUREWAY_API_KEY,
});

const { text } = await generateText({
  model: futureway("futureway-smart"),
  prompt: "Fasse den EU AI Act in drei Sätzen zusammen.",
});
console.log(text);

n8n, Make and other automation tools

Any OpenAI node that accepts a base URL works: enter URL and key, pick a model alias. For tools without a base URL field, use an HTTP request node against /chat/completions.