NeuForge Pay Python SDK (neuforge-pay)

Real-Time AI Token Metering, Margin-Aware COGS Tracking, and Usage Billing for AI Developers & MCP Servers.

neuforge-pay allows AI developers, micro-SaaS builders, and Model Context Protocol (MCP) server authors to add real-time token metering, gross margin guardrails, and usage-based billing to any FastAPI, Node.js, or MCP server in 3 lines of code.

⚡ Quick Start

1. Installation

pip install neuforge-pay

2. Add 3-Line Middleware to Your API / MCP Server

from fastapi import FastAPI
from neuforge_pay import meter_endpoint

app = FastAPI()

@app.get("/v1/agent")
@meter_endpoint(
    price_charged_usd=0.05,        # Bill customer $0.05 per request
    model_name="claude-3-5-sonnet", # Calculate real COGS against Claude 3.5
    estimated_input_tokens=1500,
    estimated_output_tokens=600
)
async def run_ai_agent(customer_id: str):
    # Your AI Agent logic here...
    return {"status": "success", "result": "Agent completed task"}

📊 Live COGS & Margin Engine

neuforge-pay contains built-in 2026 pricing tables for:

• OpenAI: gpt-4o, gpt-4o-mini, o1, o3-mini
• Anthropic: claude-3-5-sonnet, claude-3-5-haiku, claude-3-opus
• Google Gemini: gemini-2.5-flash, gemini-2.5-pro, gemini-3.5-flash, gemini-3.5-pro
• DeepSeek: deepseek-v3, deepseek-r1

It automatically calculates: $$\text{Realtime Gross Margin %} = \frac{\text{Customer Revenue Billed} - \text{Vendor Model Cost}}{\text{Customer Revenue Billed}} \times 100$$

If a heavy user's LLM loop causes their account gross margin to drop below your threshold (e.g., <35%), neuforge-pay triggers an instant guardrail alert to prevent margin collapse.

🧪 Run Demo Simulation

python3 examples/fastapi_mcp_server.py