2024-09-25 12:53:02 : Google released a new version of its Gemini 1.5 Pro artificial intelligence (AI) model on Tuesday. The release comes just months after the Mountain View-based tech giant released the last version of Gemini, which increased the context window to 2 million tokens. The company says these AI models, dubbed Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002, will not only deliver higher output and lower cost, but will also provide users with higher rates limit. Additionally, filter settings have been updated to make AI models more strictly adhere to instructions.
New Gemini AI model released
The company detailed the Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002 AI models in a blog post. These models are currently available as experimental model versions and are built on Gemini 1.5 Pro, which was first announced at Google I/O in May. Currently, the products are available to the company’s developers and enterprise customers. Developers can access it for free from Google AI Studio and Gemini API. Enterprises can access it through Vertex AI.
The latest Gemini 1.5 Pro and Flash models also perform better than the previous generation Gemini models, according to internal testing conducted by Google. The company claims that the new model achieves a 7% improvement on the Massive Multi-Task Language Understanding Pro (MMLU-Pro) benchmark. Additionally, the AI model is said to have improved by approximately 20% on the MATH and HiddenMath benchmarks compared to Gemini 1.5 Pro.
Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002 AI models will also offer higher rate limits. Rate limits are daily usage limits for users. For the 1.5 Flash model, users will get 2,000 requests per minute (RPM), while the 1.5 Pro model will offer 1,000 RPM. Google says these limits are being increased to allow developers to build with new versions of Gemini.
It’s not just rate limiting that’s being upgraded. With Gemini-1.5-Pro-002 and Flash-002, the company has also increased the number of output tokens per second, making the model more responsive and generating long blocks of text faster.
A key upgrade to these AI models is filter improvements. Google says that thanks to these updated filters, the new Gemini AI models will better obey prompts and follow instructions. Google is also improving its set of security measures to ensure that AI models don’t produce anything harmful. It’s worth noting that default filters will not be applied to new AI models, allowing developers to choose their preferred configuration.
Follow us On Social Media Twitter/X
