Google’s Pixel Recorder now uses Gemini Nano with Multimodality

Using Gemini Nano with Multimodality Recorder app now allows image and audio input

Tech Desk - Sep 02, 2024

An undated image of Google Pixel smartphone. — Unsplash

Google’s Pixel Recorder app has received a new upgrade, Gemini Nano with Multimodality, on the Pixel 9 series similar to the Pixel Screenshots app.

Pixel Recorder allows users to record, transcribe, save, and share audio making it easier for them to manage and revisit their recordings.

Previously, the Recorder app was integrated with Gemini Nano, a powerful on-device large language model (LLM). This integration introduced an AI-powered audio summarisation feature that helps users find the right recordings and quickly summarise the key points.

However, last month, Gemini Nano got a power boost by introducing the new Gemini Nano with a Multimodality model.

Now, on the Pixel 9 series, the Recorder app is using this model which allows image and audio input in addition to text. It has an expanded capability that lets the Recorder app summarise much longer transcripts than before.

It appears that this new model is significantly larger, more capable, and more accurate than the previous one on Pixel 8 devices.

Furthermore, since the integration of Gemini Nano, Recorder app users have been using the new AI-powered summarisation feature averaging 2 to 5 times daily and the number of overall saved recordings has increased by 24%, contributing to a significant increase in app engagement overall.