Damian Brady from Microsoft Developer demonstrates how developers can leverage GitHub Models to compare more than 40 AI coding models. This video helps developers select the best AI solution for their workflow.

Evaluating AI Models for Coding with GitHub Models

Presented by Damian Brady from Microsoft Developer, this video explains how the GitHub Models tool empowers developers to make informed decisions when integrating AI into their coding workflows.

What is GitHub Models?

GitHub Models allows developers to:

  • Test and compare 40+ AI models side-by-side using their own prompts
  • Evaluate output quality across various coding tasks (e.g., code completion, function generation, refactoring)
  • Avoid trial-and-error guessing when deciding which AI solution is most effective for a specific need

Key Features

  • Model Comparison: Easily try out different models and view their responses to the same prompt.
  • Prompt Engineering Guidance: See how varying your prompts can affect results, supporting prompt optimization.
  • Developer Focused: Designed for hands-on technical evaluation, rather than marketing claims or generic summaries.

Getting Started

  • Access GitHub Models via the official GitHub Models website (link from the video description)
  • Craft your own prompts to assess:
    • Code generation
    • Bug fixing
    • Documentation suggestions
    • Framework-specific tasks
  • Quickly identify which model best suits project requirements

Practical Use Cases

  • Selecting the optimal AI model for a new codebase
  • Comparing proprietary and open-source models for efficiency and accuracy
  • Learning prompt engineering techniques that improve code outputs

Additional Resources

About the Presenter

Damian Brady is a developer advocate with Microsoft, specializing in developer tooling and AI integrations.