For now we are talking about a beta version. But anyone can try its capabilities. The launch of Gemini 1.5 Pro for everyone was announced by Jeff Dean from Google DeepMind. According to him, application developers can evaluate the capabilities of the neural network in the AI Studio user interface. But there is a limit of up to 20 requests per day.
Gemini 1.5 Pro context window is 128,000. But it can be increased up to 1,000,000 tokens (and an option of 10,000,000 is being tested). At a time, the neural network processes up to 60 minutes of video, 11 hours of audio, 30,000 lines of code and about 700,000 words. Gemini 1.5 Pro is based on the Transformer and MoE architecture and is capable of quickly and accurately analyzing documents and finding information in them. According to the results of the Needle In A Haystack test, the neural network finds the necessary data in large files with an accuracy of 99%. It is also one of the leaders in the speed of adaptive learning.
Gemini 1.5 Pro context window is 128,000. But it can be increased up to 1,000,000 tokens (and an option of 10,000,000 is being tested). At a time, the neural network processes up to 60 minutes of video, 11 hours of audio, 30,000 lines of code and about 700,000 words. Gemini 1.5 Pro is based on the Transformer and MoE architecture and is capable of quickly and accurately analyzing documents and finding information in them. According to the results of the Needle In A Haystack test, the neural network finds the necessary data in large files with an accuracy of 99%. It is also one of the leaders in the speed of adaptive learning.