Migrating from OpenAI

If you already have an integration built on the OpenAI or Anthropic SDK, moving to MyTokenGate takes three changes: point the base URL at the MyTokenGate gateway, swap in your MyTokenGate API key, and use a MyTokenGate model id. The request and response formats stay the same, so the rest of your code keeps working.

Migrating an OpenAI integration

The gateway is OpenAI-compatible. Keep the openai SDK you already use and change only the base URL, the key, and the model.

Before:

from openai import OpenAI
 
client = OpenAI(
    api_key="sk-your-openai-key",
)
 
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

After:

from openai import OpenAI
 
client = OpenAI(
    base_url="https://gateway.mytokengate.com/v1",
    api_key="tg-your-api-key",
)
 
response = client.chat.completions.create(
    model="gpt-5.4",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

The SDK appends /chat/completions to the base URL, so you set the base URL to https://gateway.mytokengate.com/v1 and nothing more.

The same request as a raw curl call:

curl https://gateway.mytokengate.com/v1/chat/completions \
  -H "Authorization: Bearer tg-your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.4",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Node.js is the same idea: pass baseURL: "https://gateway.mytokengate.com/v1" and your tg- key to the openai client, then change the model.

Migrating a Claude / Anthropic integration

The gateway is also Anthropic-compatible for Claude models. Keep the anthropic SDK and change the base URL, the key, and the model.

Before:

from anthropic import Anthropic
 
client = Anthropic(
    api_key="sk-ant-your-key",
)
 
message = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}],
)
print(message.content[0].text)

After:

from anthropic import Anthropic
 
client = Anthropic(
    base_url="https://gateway.mytokengate.com/v1",
    api_key="tg-your-api-key",
)
 
message = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}],
)
print(message.content[0].text)

The SDK appends /messages to the base URL, so the base URL stays https://gateway.mytokengate.com/v1.

For Claude models, use the native Anthropic protocol shown above rather than calling them through the OpenAI-style endpoint. The native path passes features like extended thinking through directly and avoids protocol conversion, so it gives the best compatibility. Cross-protocol calls work, but native is the recommended default.

Model name mapping

Change your old model names to the current MyTokenGate ids.

This is a short list of common cases. For the full set of active models, ids, and pricing, see the Models Reference.

What stays the same

Because the gateway is protocol-compatible, everything else works the way it does in the original API:

• Request and response shapes are unchanged.
• Streaming works the same way.
• Function calling and tool use work the same way.
• JSON mode and structured output work the same way.

You do not rewrite your handling code. Point the client at the gateway and keep parsing responses as before.

Common pitfalls

• Do not include /v1 twice. Set the base URL to https://gateway.mytokengate.com/v1 and let the SDK append /chat/completions or /messages. Adding the path yourself produces a doubled route.
• Keep the native protocol per model family. Use the OpenAI protocol for GPT, Gemini, and other models, and the Anthropic protocol for Claude.
• Use your MyTokenGate tg- key, not the vendor key. An old sk- or sk-ant- key will not authenticate against the gateway.

Next steps

• Chat Completions reference
• Tools Integration