Usage Guide
Install, configure, run evaluations, compare models, and use Python API.
Open Usage GuideAegis Monitor Documentation
LLM evaluation, cost monitoring, regression detection, and model comparison.
Use this page as the documentation homepage for the library.
Full Documentation
Complete project reference across architecture, workflows, and operations.
Open Full DocumentationArchitecture
Detailed system design, module boundaries, extension points, and data flow.
Open ArchitectureAdapter Development
Create and test custom provider adapters with implementation guidance.
Open Adapter DevelopmentContributing
Project standards, PR process, testing expectations, and setup steps.
Open ContributingCode of Conduct
Community behavior expectations and reporting process.
Open Code of ConductQuick Start
pip install aegis-ai
# evaluate a dataset
aegis eval run --dataset examples/datasets/qa_sample.yaml --model gpt-4
# compare multiple models
aegis compare --dataset examples/datasets/qa_sample.yaml --models gpt-4,gpt-3.5-turbo,claude-3-opusExamples
simple_eval.py · model_compare.py · cost_tracking_demo.py · sentiment_classification.py · qa_evaluation.py