Google AI Studio Product Director Logan Kilpatrick tweeted on X (August 28),Announced the launch of 3 Gemini Experimental Model.
The three experimental Gemini AI models launched by Google this time are as follows:
Gemini 1.5 Flash-8B
Gemini 1.5 Flash-8B is a smaller model of Gemini 1.5 Flash with 8 billion parameters, designed for multimodal tasks, including high-capacity tasks and long text summarization tasks.
Gemini 1.5 Pro Exp-0827
The main enhancement, programming, complex prompt words, is now available for free through Google AI Studio and Gemini API as "gemini-1.5-pro-exp-0827".
Kilpatrick said the new Gemini 1.5 Pro Exp 0827 model is superior in every way to the experimental model released in early August and is currently ranked No. 2 on LMSYS, just behind OpenAI’s GPT-4o-latest model.
Starting September 3, Google will automatically redirect requests for the gemini-1.5-pro-exp-0801 model to the new gemini-1.5-pro-exp-0827 model.
The gemini-1.5-pro-exp-0801 model will be removed from Google AI Studio and the API.
Gemini 1.5 Flash Exp-0827
The Gemini-1.5-Flash (0827) version has significant performance improvements, and its ranking on LMSYS has risen from 23rd to 6th.
Users can access the above two models through Gemini API and Google AI Studio, named gemini-1.5-pro-exp-0827 and gemini-1.5-flash-exp-0827 respectively.