Case Study: Kilo Code stops AI and LLM abuse with Stytch Device Fingerprinting

Kilo Code stops AI abuse and serves 100B tokens a day with Stytch

Kilo Code, an open-source AI coding assistant, faced a significant challenge with automated abuse after launching a free credit strategy to attract users. Attackers created thousands of fake accounts to claim these credits, threatening to divert their growth budget. Traditional defenses like CAPTCHAs were ineffective, and the company needed a solution that would stop abuse without adding friction for legitimate developers. They turned to Stytch for help.

By integrating Stytch Device Fingerprinting, Kilo Code could silently assess the risk of each signup. Suspicious users were required to verify a credit card to receive credits, while real users experienced a frictionless process. This solution from Stytch allowed Kilo Code to safely scale its growth, climb to the #1 AI coding tool on OpenRouter, and serve over 100 billion tokens daily, all while freeing the team to focus on product development instead of building abuse detection.

Kilo Code