Google has introduced several Gemini models before, and at the Google I/O 2024 event today, Google also announced Gemini 1.5 Flash. As the name suggests, this model focuses on generating capabilities that work faster than other Gemini models.
Gemini 1.5 Flash is aimed at fast generation support that doesn't require a lot of analysis. It can be used by developers in developing a number of things, including for customer service response, image generation, and so on. The Pro or Ultra models are seen to target more in-depth and precise analysis needs.
This Gemini 1.5 Flash is said to support up to 1 million tokens through it. Gemini Flash also offers multi-modal support like other Gemini models – simultaneously supporting not only text, but also images and videos in terms of input or output.
With this introduction, Gemini is now offered in the form of "Nano", "Flash", "Pro" and also the most powerful is "Ultra".
Google also shared that Gemini is now used by more than 1.5 million developers around the world in developing their respective applications. Also in line with today's announcement, Gemini 1.5 Pro is accessible to users through Gemini Advanced in 35 different languages.
At the same time, despite already supporting the generation of more than 1 million tokens, Google also moved forward and stated that it is offering the ability of up to 2 million tokens in a limited form to a few selected developers. With this higher token support, it simultaneously allows Gemini to answer more things, including to formulate a 1500-page book.
In addition to developers, the new Gemini model will also be widely integrated into existing Google products. For example, on GMail, Gemini can assist in summarizing and summarizing various emails and the documents attached to them.
Moving forward, Gemini looks to be fully integrated into Google products – not just search, and will also replace Google Assistant on Android phones.