OpenAI Compatibility

GoingMerry provides a built-in OpenAI-compatible API gateway. You can use standard OpenAI SDK clients, LangChain modules, or web frontends by redirecting request traffic to your local GoingMerry daemon.

• Base URL: http://localhost:11434/v1
• API Key: Any dummy string value (e.g. merry or ignored).

---

Supported Endpoints

• POST /v1/chat/completions -> Mapped directly to /api/chat
• POST /v1/embeddings -> Mapped directly to /api/embeddings

---

Python Integration

Use the official openai Python package by modifying the client parameters:

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:11434/v1",
    api_key="merry",  # Required but ignored by GoingMerry
)

response = client.chat.completions.create(
    model="gemma4",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the primary optimization in GoingMerry?"}
    ],
    temperature=0.7,
    stream=True
)

for chunk in response:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

---

Node.js Integration

Modify the standard configuration using the Javascript/TypeScript library:

import OpenAI from 'openai';

const openai = new OpenAI({
  baseURL: 'http://localhost:11434/v1',
  apiKey: 'merry', // Required but ignored
});

async function main() {
  const stream = await openai.chat.completions.create({
    model: 'gemma4',
    messages: [{ role: 'user', content: 'Explain branchless programming' }],
    stream: true,
  });

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

main();

---

LangChain Setup

Redirect completion calls inside LangChain python:

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="gemma4",
    openai_api_key="merry",
    openai_api_base="http://localhost:11434/v1"
)

response = llm.invoke("Hello LangChain!")
print(response.content)

---

LiteLLM CLI Execution

Test routing using the LiteLLM proxy boundary:

litellm --model openai/gemma4 --api_base http://localhost:11434/v1