Google Unveils Gemini 1.5 Pro With Record 1 Million Token Context Window
Google launches a private preview of Gemini 1.5 Pro in Google AI Studio, featuring a groundbreaking experimental 1 million token context window. The new model utilizes a Mixture-of-Experts architecture to deliver faster and higher quality responses.
Google introduces Gemini 1.5 Pro, its next-generation model now available for private preview in Google AI Studio. This mid-sized multimodal model utilizes a new Mixture-of-Experts (MoE) architecture that routes requests to smaller expert neural networks, resulting in faster and higher quality responses. Developers eager to test this advanced technology sign up through Google AI Studio, which serves as the fastest way to build with and integrate Gemini API models.
The standout feature of Gemini 1.5 Pro is its experimental 1 million token context window, a massive leap from the previous industry benchmark of 200,000 tokens. While the standard model includes a 128,000 token context window, private preview participants get access to this unprecedented capacity. This expanded window unlocks entirely new use cases by allowing the model to process vast amounts of information in a single prompt.
With this massive context window, developers directly upload entire code repositories, lengthy videos, or multiple large PDFs directly into Google AI Studio as prompts. The model then reasons across these different modalities to output comprehensive text answers. By taking in significantly more information at once, Gemini 1.5 Pro delivers much more consistent and reliable outputs for complex developer queries.