Welcome to **Selene API**, your go-to solution for evaluating generative AI applications. With our advanced models, you can easily find and fix AI mistakes at scale, ensuring your GenAI applications are reliable and effective.

What are the features of Selene API?

* **State-of-the-art models**: Selene outperforms other models on evaluation benchmarks, making it the most accurate choice for your needs. * **Customizable evaluations**: Tailor your evals to fit your specific use case with our Eval Copilot (beta), allowing for fine-grained scoring and criteria. * **Accurate scores and critiques**: Integrate our API seamlessly into your workflow to generate precise eval scores along with actionable feedback. * **Optimized for speed**: Selene 1 Mini is the best evaluation model of its size, perfect for running evals quickly during inference time.

What are the use cases of Selene API?

* Detecting hallucinations in RAG applications. * Comparing AI-generated answers to a ground truth. * Evaluating the relevance, correctness, and helpfulness of AI responses.

How to use Selene API?

To get started, sign up for a free account and explore our API documentation. Use the Eval Copilot (beta) to customize your evaluation metrics and run your first eval with Selene models.