Framework Integrations

The Curate-Me gateway works with any framework that uses OpenAI-compatible or Anthropic-compatible APIs. Swap the base URL and every LLM call gets governance — no SDK changes.

LangChain / LangGraph


from langchain_openai import ChatOpenAI
 
llm = ChatOpenAI(
    model="gpt-4o",
    base_url="https://api.curate-me.ai/v1/openai",  # One line change
    default_headers={"X-CM-API-Key": "YOUR_GATEWAY_KEY"},
)
 
response = llm.invoke("Explain quantum computing")

Every chain and agent call now flows through the gateway with cost tracking, rate limiting, and PII scanning.

CrewAI

CrewAI reads the OPENAI_BASE_URL environment variable:


export OPENAI_BASE_URL=https://api.curate-me.ai/v1/openai
export OPENAI_API_KEY=YOUR_PROVIDER_KEY
export OPENAI_DEFAULT_HEADERS='{"X-CM-API-Key": "YOUR_GATEWAY_KEY"}'


from crewai import Agent, Task, Crew
 
researcher = Agent(role="Researcher", goal="Find data", llm="gpt-4o")
crew = Crew(agents=[researcher], tasks=[Task(description="Research AI trends")])
crew.kickoff()  # Governed by Curate-Me

OpenAI Agents SDK


from openai import OpenAI
 
client = OpenAI(
    base_url="https://api.curate-me.ai/v1/openai",  # One line change
    api_key="YOUR_PROVIDER_KEY",
    default_headers={"X-CM-API-Key": "YOUR_GATEWAY_KEY"},
)
 
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)

Claude / Anthropic SDK


import anthropic
 
client = anthropic.Anthropic(
    base_url="https://api.curate-me.ai/v1/anthropic",  # One line change
    api_key="YOUR_ANTHROPIC_KEY",
    default_headers={"X-CM-API-Key": "YOUR_GATEWAY_KEY"},
)
 
message = client.messages.create(
    model="claude-sonnet-4-5-20250929",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}],
)

TypeScript / Node.js


import OpenAI from 'openai'
 
const client = new OpenAI({
  baseURL: 'https://api.curate-me.ai/v1/openai',
  apiKey: 'YOUR_PROVIDER_KEY',
  defaultHeaders: { 'X-CM-API-Key': 'YOUR_GATEWAY_KEY' },
})
 
const response = await client.chat.completions.create({
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Hello!' }],
})

Verifying Governance

After making a request, check that governance is active:


# Check your dashboard for the request
curl -s https://api.curate-me.ai/v1/models \
  -H "X-CM-API-Key: YOUR_GATEWAY_KEY" | jq '.data | length'
# Should return the number of available models

Visit dashboard.curate-me.ai/costs to see the request logged with cost, model, and governance status.

What You Get

Every request through the gateway automatically receives:

Feature	What happens
Cost tracking	Request cost recorded in real-time (Redis + MongoDB)
Rate limiting	Per-org, per-key request throttling
PII scanning	Regex scan for secrets/PII before hitting the provider
Model allowlist	Only approved models per your org policy
Budget caps	Daily/monthly spend limits enforced
HITL approval	High-cost requests flagged for human review
Audit trail	Every request logged with full metadata