Tool setup

Any OpenAI-compatible tool

The generic recipe for any SDK or client that speaks the OpenAI API.

Anything that speaks the OpenAI API — the official SDKs, LangChain, LlamaIndex, Aider, Continue, your own scripts — works with the gateway. This is the generic recipe.

What you get

  • Any OpenAI-compatible client pointed at the gateway, billed to your project's sponsor budget, with no code changes beyond the base URL and key.

Prerequisites

  • Your project access key and base URL from your project control panel — YOUR_PROJECT_KEY and https://v1.opentokenmarket.dev.

The two values

SettingValue
Base URL (OpenAI-style clients)https://v1.opentokenmarket.dev
Base URL (Anthropic-style clients)https://v1.opentokenmarket.dev
API key / bearer tokenYOUR_PROJECT_KEY
Modelany Model name from the model catalog

Most tools read these from OPENAI_BASE_URL and OPENAI_API_KEY:

export OPENAI_BASE_URL="https://v1.opentokenmarket.dev"
export OPENAI_API_KEY="YOUR_PROJECT_KEY"

Examples

OpenAI SDK (Python)

from openai import OpenAI

client = OpenAI(
    base_url="https://v1.opentokenmarket.dev",
    api_key="YOUR_PROJECT_KEY",
)

resp = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Say hi in 3 words."}],
)
print(resp.choices[0].message.content)

OpenAI SDK (Node)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://v1.opentokenmarket.dev",
  apiKey: process.env.OPENAI_API_KEY,
});

const resp = await client.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Say hi in 3 words." }],
});
console.log(resp.choices[0].message.content);

cURL

curl https://v1.opentokenmarket.dev/chat/completions \
  -H "Authorization: Bearer YOUR_PROJECT_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o","messages":[{"role":"user","content":"ping"}]}'

Anthropic-style clients

The gateway also serves the Anthropic Messages API. Point ANTHROPIC_BASE_URL at https://v1.opentokenmarket.dev (no /v1) and send the key as x-api-key or ANTHROPIC_AUTH_TOKEN. See the Claude Code guide for a worked example.

Troubleshooting

  • 404 on /chat/completions — your base URL is missing the trailing /v1.
  • 401 — the key isn't your project access key, or it's sent in the wrong header. Use Authorization: Bearer <key> for OpenAI-style, x-api-key: <key> for Anthropic-style.
  • model not found — use an exact name from the model catalog.
  • budget has been exceeded — the project's funded balance is spent. Ask your sponsor to top up.

List models programmatically

curl https://v1.opentokenmarket.dev/models -H "Authorization: Bearer YOUR_PROJECT_KEY"