Back to all discussions

  •   over 1 year ago

Using Combination of Build-in AI + Gemini AI

Hi everyone!

The built-in summarizer and prompt APIs have relatively small context sizes and it will be very hard for me to accomplish my goals. In result I will need to use Gemini 1.5 Flash-8B or GEMMA... My extension will not rely only on external models, it will be using the language detection API too, is this allowed?

P.S. The Overview section says "The APIs include but are not limited to the Prompt API for Web and in Chrome Extensions, Summarization API, Write API, and Rewrite API".

  • 5 comments

  • Manager   •   over 1 year ago

    If you're looking to create a hybrid AI application with both server-side and built-in, that would work. But, your submission should not fully be relying on Gemini 1.5. If you need further clarification, feel free to join our Discord, where someone a bit more technical than me can also help: https://discord.com/invite/google-dev-community

  •   •   over 1 year ago

    Thanks for the answer! Just one more thing to clarify as I think many people ask themselves the same: Are hybrid AI extension allowed in this case? Because extensions can communicate with a backend...

  • Manager   •   over 1 year ago

    Yes, they are allowed and encouraged, and we will be on the lookout for Hybrid AI Extensions for those as one of our winners!

  •   •   over 1 year ago

    So, Projects using hybrid AI will be eligible for 'Best Hybrid AI App/Extension' but not the main prizes. Am i right?

  • Manager   •   over 1 year ago

    No, a hybrid AI application can also be eligible for one of the other prizes (e.g. 'Best Real-World' or 'Most Innovative'.) We created an additional 'Best Hybrid AP App/Extension' category to encourage these types of submissions.

Comments are closed.