Local REST API

The Privane CLI bundles a highly efficient REST server that matches the OpenAI API specification. This allows you to use Privane as a drop-in replacement for any tool that expects OpenAI endpoints (like LangChain, Cursor IDE, or AutoGen).

OpenAI-Compatible by Design

Exposing an OpenAI-compatible API gateway allows existing AI frameworks, development environments, and software agents to run fully locally with exactly zero code changes.

This completely eliminates vendor lock-in and allows you to transition your existing cloud-dependent apps to a private, localized ecosystem overnight.

Privane functions as a seamless drop-in replacement for:

• Orchestration Frameworks: LangChain, AutoGen, CrewAI, LlamaIndex
• Agentic Ecosystems: OpenHands, AutoGPT
• AI Coding Interfaces: Cursor, Continue.dev, VSCode extensions
• Standard Clients: Official OpenAI SDKs (Python, Node.js, Go, Rust)

Starting the Server

privane serve

This starts a local instance running on http://localhost:8080.

Example: cURL

curl http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemma-2b",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Explain quantum computing in simple terms."
      }
    ]
  }'

Example: LangChain

You can easily configure the official LangChain OpenAI wrappers to point to your local Privane instance.

from langchain_openai import ChatOpenAI
 
# Point the base_url to the Privane Local API
llm = ChatOpenAI(
    base_url="http://localhost:8080/v1",
    api_key="not-needed",
    model="gemma-2b"
)
 
response = llm.invoke("What is the capital of France?")
print(response.content)