Find & Fix AI Chat Flaws Before You Ship
Open-source framework for stress-testing LLMs and conversational AI. Identify hallucinations, policy violations, and edge cases with scalable, realistic simulations.
A Toolkit Built for Production-Ready AI
Stop manually testing your chatbot. Automate the process of finding critical failures with a framework built for modern development workflows.
Simulate Realistic Conversations at Scale
Run thousands of parallel conversations using diverse user personas. Test your AI's performance, tone, and logic under real-world pressure.
Uncover Critical Risks & Edge Cases
Proactively detect hallucinations, toxic output, jailbreak prompts, and logical failures before they impact your users.
Why Build with OneRun?
Ready to build more reliable AI? Choose your path below.
Full Control, No Black Boxes
Tired of SaaS tools with opaque policies and data privacy risks? OneRun is open source and self-hostable. Run it in your own VPC, on-prem, or locally. You control the environment, the data, and the code.
Simulate Thousands Of Scenarios
Set up a complete testing environment with a single docker-compose up command. Define personas and test cases in simple YAML files and launch thousands of simulations from the command line.
Free & Community-Driven
OneRun is and always will be free. Built by developers for developers, our roadmap is driven by community contributions. No vendor lock-in, no surprise billing.
< Get started >