Google has recently announced an upgrade to its AI chatbot Gemini. Gemini 1.5 boasts significant advancements in terms of performance and efficiency. This new version builds upon the success of its predecessor which was rolled out last week. The latest update incorporates innovations across various aspects of model development and infrastructure. The introduction of a new Mixture-of-Experts (MoE) architecture into it has made it more efficient to train and serve.


The initial release of Gemini 1.5 includes Gemini 1.5 Pro which is a mid-size multimodal model. This model has been optimised for scalability across diverse tasks. Despite the smaller size, Gemini 1.5 Pro delivers performance comparable to 1.0 Ultra. Not only that, it also introduces breakthrough features for long-context understanding.


Gemini 1.5 Pro offers a 128,000-token context window, with a limited preview which allows developers and enterprise customers to experiment with up to a million of tokens. This expanded context window enables the model to process a huge amount of information, including text, audio, videos, and codebases.


One of the key features of Gemini 1.5 is learning complex tasks more quickly and maintaining quality while increasing efficiency in training and serving.


The enhanced context window capacity allows for more consistent, relevant, and useful output by processing more amount of information in a prompt.


What Can Gemini 1.5 Do?


Gemini 1.5 Pro can analyse, classify, and summarise large amounts of content within a given prompt which ultimately enables it to understand and reason about complex topics. It can perform sophisticated understanding and reasoning tasks across different modalities, including video, which enhances its capabilities in analysing diverse types of data.


It excels at problem-solving tasks across longer blocks of code. It provides helpful solutions, modifications, as well as explanations.


Gemini 1.5 Pro outperforms its predecessor on a number of benchmarks.


Availability & Pricing


A limited preview of Gemini 1.5 Pro is available to developers and enterprise customers via AI Studio and Vertex AI. As per reports, Google is planning to introduce pricing tiers based on context window size, with early testers able to experiment with the 1 million token context window at no cost during the testing period.