Scorecards and Reports Test results are encrypted, stored and displayed locally on your machine to ensure authenticity and protect any sensitive information. Scorecards are designed to render past and current test runs, and to be easily understood and shared. Each report includes a summary of the test results, associated benchmark comparisons, and an opportunity to dig deeper and review the test result details.
See more on the scoring methodology in the documentation published in the
LLM Canary Github Repo.Watch this quick demo to learn more about the scorecards in action.
LLM Canary Benchmarks LLM Canary's benchmarking process is designed to assess and compare various Language Learning Models (LLMs) against a basket of stable, well-known LLMs. This approach provides a consistent and reliable benchmark for evaluating the security of different models.
Review our detailed benchmarking methodology
published in the LLM Canary Github Repo 
