- Run series of tests to verify the compatibility of more recent LLMs for each providers (GPT-5, Claude-Sonnet-4 series) - Update the default models - Update the models used in the examples - Update the models in the documentation guides