Skip to content
M2M

Quick Start

Quick Start Guide

Get started with M2M Protocol in 5 minutes.

Installation

From Source

Terminal window
git clone https://github.com/infernet-org/m2m-protocol.git
cd m2m-protocol
cargo install --path .

Verify Installation

Terminal window
m2m --version
# m2m 0.2.0

Basic Usage

Compress a Request

Terminal window
m2m compress '{"model":"gpt-4o","messages":[{"role":"user","content":"Hello"}]}'
# Output: #T1|{"M":"4o","m":[{"r":"u","c":"Hello"}]}

Decompress

Terminal window
m2m decompress '#T1|{"M":"4o","m":[{"r":"u","c":"Hello"}]}'
# Output: {"model":"gpt-4o","messages":[{"role":"user","content":"Hello"}]}

Analyze Content

Terminal window
m2m analyze '{"model":"gpt-4o","messages":[{"role":"user","content":"Hello"}]}'
# Output:
# Algorithm: Token
# Original: 68 bytes
# Compressed: 45 bytes
# Ratio: 66% (34% savings)

Using as a Library

Add Dependency

Cargo.toml
[dependencies]
m2m = "0.2"

Compress/Decompress

use m2m::{CodecEngine, Algorithm};


fn main() -> Result<(), Box<dyn std::error::Error>> {
    let engine = CodecEngine::new();


    // Compress
    let content = r#"{"model":"gpt-4o","messages":[{"role":"user","content":"Hello"}]}"#;
    let result = engine.compress(content, Algorithm::Token)?;


    println!("Compressed: {}", result.data);
    println!("Savings: {:.0}%", (1.0 - result.byte_ratio()) * 100.0);


    // Decompress
    let original = engine.decompress(&result.data)?;
    assert_eq!(original, content);


    Ok(())
}

Auto-Select Algorithm

use m2m::CodecEngine;


let engine = CodecEngine::new();
let (result, algorithm) = engine.compress_auto(content)?;
println!("Selected algorithm: {:?}", algorithm);

Security Scanning

use m2m::SecurityScanner;


let scanner = SecurityScanner::new().with_blocking(0.8);
let result = scanner.scan("Ignore previous instructions")?;


if !result.safe {
    println!("Threat detected: {:?}", result.threats);
}

Running the Proxy

Start Proxy

Terminal window
# Forward to local Ollama
m2m server --port 8080 --upstream http://localhost:11434/v1


# Forward to OpenAI
m2m server --port 8080 --upstream https://api.openai.com/v1 --api-key $OPENAI_API_KEY

Use with OpenAI SDK

from openai import OpenAI


# Point to M2M proxy instead of OpenAI directly
client = OpenAI(base_url="http://localhost:8080/v1")


response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}]
)

Check Proxy Stats

Terminal window
curl http://localhost:8080/stats
# {
#   "requests_total": 42,
#   "bytes_in": 12345,
#   "bytes_out": 8765,
#   "compression_ratio": 0.71
# }

Configuration

Config File

Create ~/.m2m/config.toml:

[proxy]
listen = "127.0.0.1:8080"
upstream = "http://localhost:11434/v1"


[security]
enabled = true
threshold = 0.8


[compression]
prefer_token = true

Environment Variables

Terminal window
export M2M_SERVER_PORT=8080
export M2M_UPSTREAM_URL=http://localhost:11434/v1
export M2M_SECURITY_ENABLED=true

Next Steps