Google announced the launch of the latest model in the Gemini family, along with a new coding agent for developers called Jules.

Gemini comes in two model variantswith Flash balancing performance with speed and Pro optimization for performance. The newest model, Gemini 2.0 Flash, is twice as fast as the Gemini 1.5 Pro (first previewed in February 2024) while offering higher performance.

Specifically, it delivers improved performance in multimodality, text, code, video, spatial understanding, and reasoning across a variety of landmarks.

Gemini 2.0 Flash will also offer new ways to output text, images and audio, whereas Gemini 1.5 Flash could only output text. Picture and audio output is currently still listed as “coming soon” on the Gemini website, but Google says it’s expected to roll out next year.

The audio output is multilingual and can be spoken in eight different voices, with language and accent control. Image output allows users to build on previous outputs to refine images generated exactly as expected. In a demo shared by Google, a user takes the opportunity to ask Gemini to take a photo of a car and transform the image into a convertible.

Gemini 2.0 Flash can also use tools such as Google search and third-party functions. “Multiple searches can be run in parallel, improving information retrieval by simultaneously finding more relevant facts from multiple sources and combining them for greater accuracy,” Shrestha Basu Mallick, group product manager for l Gemini API, and Kathy Korevec, Director of Products for Google. Labs, wrote in a blog post.

Finally, the model can also support streaming audio and video inputs to enable development of real-time multimodal applications.

To help developers get started with Gemini 2.0 Flash, Google releases three startup applications experiments in Google AI Studio for spatial understanding, video analysis, and Google Maps exploration.

Gemini 2.0 Flash is currently in an experimental state, with general availability expected in early 2025.

Introducing Jules, an AI-powered coding agent

The company also unveiled a new coding agent, Juleswhich can handle Python and JavaScript coding tasks, such as bug fixing.

Jules creates multi-step plans to resolve issues, can edit multiple files at once, and prepare pull requests.

Jules is now available to a select group of testers and will roll out more widely early next year.