⚡ Real use cases

run.pay for every
AI agent stack

From LangChain safety monitoring to CrewAI compliance — here's how developers use run.pay to give their agents superpowers.

🦜 LangChain

LLM Output Safety

Intercept every LLM response with a hallucination check before it reaches your users. One tool definition, zero setup.

Hallucination Detector PII Scanner Bias Detector

from langchain.tools import tool from runpay import RunPay client = RunPay(agent_id="agt_your_agent") @tool def safety_check(response: str) -> dict: """Check LLM output for hallucinations and bias.""" return client.call("hallucination-detector", { "response": response })

Try it → Full guide

🚀 CrewAI

Compliance Automation

Add a compliance agent to your crew that automatically checks every output for GDPR violations and PII before storing or sending.

GDPR Checker PII Scanner AI Act Compliance

from crewai.tools import BaseTool from runpay import RunPay client = RunPay(agent_id="agt_your_agent") class GDPRTool(BaseTool): name = "GDPR Compliance Check" description = "Check text for GDPR violations" def _run(self, text: str) -> str: result = client.call("gdprcheck", {"text": text}) return f"GDPR: {result['gdpr_compliant']}"

Try it → Full guide

🤖 AutoGen

Multi-agent Reasoning

Give your AutoGen agents access to moral reasoning, logical fallacy detection, and argument extraction for complex decision-making.

Moral Reasoning Fallacy Detector Goal Decomposer

import autogen from runpay import RunPay client = RunPay(agent_id="agt_your_agent") def moral_check(scenario: str) -> dict: """Evaluate ethical dimensions of a decision.""" return client.call("moralreason", { "scenario": scenario }) autogen.register_function(moral_check, caller=assistant)

Try it → Full guide

📊 Data pipelines

Autonomous Data Processing

Your agent generates synthetic training data, validates datasets, runs statistical analysis — paying only for what it uses.

Synthetic Data Gen CSV Validator Statistics Calculator

from runpay import RunPay client = RunPay(agent_id="agt_your_agent") # Generate 50 fake users for testing data = client.call("synthdata", { "count": 50, "schema": { "name": "name", "email": "email", "age": "integer(18,65)" } }) # $0.01 charged · 50 records returned

Try it → Full guide

Ready to give your agents superpowers?

First 3 calls free. No credit card.

Open playground →

Real code. Runs in 3 lines.

Copy-paste any example below and it works immediately with your agent ID.

Hallucination Detection

import runpay

runpay.configure(agent_id="agt_your_id")  # getrunpay.com/playground

result = runpay.call("halludetect", {
    "response": "According to Harvard, 73.2% of AI models hallucinate daily."
})

print(result["hallucination_score"])  # 87
print(result["risk"])                 # "high"
print(result["flags"])                # ["suspicious_precision", "no_source"]
print(result["_meta"]["cost"])        # 0.01

PII Scanner

result = runpay.call("piiscan", {
    "text": "Contact John at john.doe@example.com or call +1-555-0123"
})

print(result["pii_found"])    # ["email", "phone"]
print(result["redacted"])     # "Contact John at [EMAIL] or call [PHONE]"
print(result["risk_level"])   # "medium"

GDPR Compliance Check

result = runpay.call("gdprcheck", {
    "text": "We collect emails and sell data to third-party advertisers without consent."
})

print(result["compliant"])     # False
print(result["violations"])    # ["no_consent", "third_party_sharing"]
print(result["severity"])      # "critical"

Try it live → Read the docs →