Quick Read AI

Quick Read AI
Quick Read AI
Prompt API response
Summarization API
Quick Read AI Settings

💡 Inspiration

Long blogs and articles can often feel overwhelming. Our extension simplifies your browsing experience by summarizing content into key points, providing quick and insightful overviews of any webpage you explore. With the help of Google Gemini Nano capability, we have made an extension that uses an in-built AI model to perform the tasks.

✔ What it does

Quick Read AI is a browser sidebar extension designed to provide an instant overview of any webpage. It extracts data from the web, processes it, and uses the Gemini Nano model to generate insightful responses and summaries.

🔑 Key features:

➡ Summarization: The extension provides users with an instant overview of any webpage with just a click. By extracting webpage data on the backend as soon as a user opens, navigates to, or reloads a page, it ensures a concise and immediate summary of the current content.

➡ Query: Users can also ask questions related to the current webpage, and the Gemini on- device AI model instantly provides accurate answers, making it easy to extract insights and clarify details.

✨ Some Other features:

➡ Accessibility

Offers both Dark and Light modes to enhance user comfort and adaptability.
Includes Speech Input functionality, enabling hands-free interaction for improved accessibility.

➡ Model Setting: The model settings can be customized to adjust the intensity and detail of the responses, tailoring the experience to individual preferences.

🛠How we built it

We have developed a sidebar extension using React and JavaScript, powered primarily by the Gemini Nano model for generating responses. As a fallback, the Gemini Flash-1.5 model ensures reliability.

➡ Leveraged the Gemini Nano Prompt API to generate responses specific to the current webpage by utilizing a system prompt.

➡ Integrated the Summarization API to create concise summaries of extracted webpage content.

➡ Implemented the Web Speech API to enable seamless speech-to-text functionality for user convenience.

⛔ Challenges we ran into

➡ We had difficulty getting the appropriate result from the Prompt API model and had multiple model failures/crashes.

➡ Dealing with content script was a bit challenging for us while implementing some features.

🚀 Accomplishments that we're proud of:

We implemented an API that tailors responses to the specific structure of each webpage. By customizing the prompts based on the unique data and layout of each page, we ensure that the responses are relevant and accurate.

😎 What we learned:

This was our first experience building a browser extension and working with a built-in model like Gemini Nano. The process was both challenging and rewarding, allowing us to explore the practical use cases of built-in AI models. We've gained valuable insights and look forward to leveraging these models in future projects and ideas as new updates are released.

⏭ What's next for Easy Browsing

Our next step is to integrate browser controls into the extension using browser-specific APIs, enhancing functionality. Additionally, we plan to improve accessibility by adding a Translation API to allow users to translate both responses and entire webpages for a more inclusive experience.