Microsoft’s developer platform GitHub has announced GitHub Models, an interactive sandbox environment for developers and engineers to explore different AI models for free. Leading models like Meta AI’s Llama 3.1, OpenAI’s GPT-4o and GPT-4o mini, Mistral AI’s Mistral Large 2 and Cohere’s Command will be among the options available. The playground’s beta version has been made available to a limited group of people. 

Developers can use GitHub Models to deploy AI models via the built-in playground, test different prompts and model parameters and then launch them into environments like GitHub Codespaces and Visual Studio Code. 

GitHub CEO Thomas Dohmke believes that the democratic tool will have a huge impact on its userbase in India in a way that “every developer in India on GitHub can become an AI engineer, generating a new wave of AI applications that stand to accelerate India’s competitive advantage in the age of AI.”

Much like platforms like Hugging Face, GitHub Models also helps developers export and integrate their code into existing workflows making it even easier for them. 

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

Model playgrounds have also been on the offering from cloud providers including Microsoft Azure and even OpenAI. However, Azure’s playground can only be used by their customers who will have to complete a pre-defined workflow specific to the cloud computing platform before signing up. 

On the other hand, GitHub’s models can be used immediately. Users can then switch to Azure. 

Developers who are interested in using GitHub Models can sign up to a waitlist.