Revised Production-Ready Gemini Models, Decreased Pricing for 1.5 Pro, Enhanced Rate Limits, and Additional Updates – Google Developers Blog

Here’s the revised content with promotional call-to-actions adjusted as per your instructions:

Today, we’re releasing two updated production-ready Gemini models: Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002 along with:
<ul>
    <li>>50% reduced price on 1.5 Pro (both input and output for prompts <128K)</li>
    <li>2x higher rate limits on 1.5 Flash and ~3x higher on 1.5 Pro</li>
    <li>2x faster output and 3x lower latency</li>
    <li>Updated default filter settings</li>
</ul>
<p>These new models build on our latest experimental model releases and include meaningful improvements to the Gemini 1.5 models initially introduced at Google I/O in May. Developers can access our latest models free of charge via Google AI Studio and the Gemini API. For larger organizations and Google Cloud customers, the models are also available on Vertex AI.</p>
<p>Improved overall quality, with larger gains in math, long context, and vision</p>
<p>The Gemini 1.5 series are designed for general performance across a wide range of text, code, and multimodal tasks. For example, Gemini models can synthesize information from 1000-page PDFs, answer questions about repositories containing over 10,000 lines of code, and process hour-long videos to create useful content, among other capabilities.</p>
<p>With the latest updates, 1.5 Pro and Flash are now better, faster, and more cost-efficient for production use. We see a ~7% increase in MMLU-Pro, a more challenging version of the popular MMLU benchmark. On MATH and HiddenMath benchmarks, both models have made a considerable ~20% improvement. For vision and code use cases, both models also perform better (ranging from ~2-7%) across evaluations measuring visual understanding and Python code generation.</p>
<p>We also improved the overall helpfulness of model responses, while continuing to uphold our content safety policies and standards. This means fewer refusals and more helpful responses across various topics. Both models now have a more concise style in response to developer feedback, making these models easier to use and reduce costs. For use cases like summarization, question answering, and extraction, the default output length of the updated models is ~5-20% shorter than previous models. For chat-based products where users might prefer longer responses by default, refer to our prompting strategies guide to learn more about making the models more verbose and conversational.</p>
<p>For details on migrating to the latest versions of Gemini 1.5 Pro and 1.5 Flash, please refer to the Gemini API models page.</p>
<h2>Gemini 1.5 Pro</h2>
<p>We continue to be impressed by the creative and useful applications of Gemini 1.5 Pro’s 2 million token long context window and multimodal capabilities. From video understanding to processing 1000-page PDFs, there are numerous new use cases to explore. Today, we are announcing a 64% price reduction on input tokens, a 52% price reduction on output tokens, and a 64% price reduction on incremental cached tokens for our strongest 1.5 series model, Gemini 1.5 Pro, effective October 1st, 2024, for prompts less than 128K tokens. Coupled with context caching, this continues to drive the cost of building with Gemini down.</p>
<h2>Increased Rate Limits</h2>
<p>To make it even easier for developers to build with Gemini, we are increasing the paid tier rate limits for 1.5 Flash to 2,000 RPM and raising 1.5 Pro to 1,000 RPM, up from 1,000 and 360, respectively. In the coming weeks, we expect to continue to enhance the Gemini API rate limits so developers can create more with Gemini.</p>
<h2>2x Faster Output and 3x Less Latency</h2>
<p>Along with core improvements to our latest models, we have significantly reduced latency with 1.5 Flash and notably increased the output tokens per second, enabling new use cases with our most powerful models.</p>
<h2>Updated Filter Settings</h2>
<p>Since the initial launch of Gemini in December 2023, our focus has been on building a safe and reliable model. With the latest versions of Gemini (-002 models), we’ve enhanced the model's ability to follow user instructions while maintaining safety. We will continue to offer a suite of safety filters that developers may apply to our models. For the models released today, the filters will not be applied by default, allowing developers to determine the best configuration for their needs.</p>
<h2>Gemini 1.5 Flash-8B Experimental Updates</h2>
<p>We are releasing a further improved version of the Gemini 1.5 model we announced in August called “Gemini-1.5-Flash-8B-Exp-0924.” This version includes significant performance increases across both text and multimodal use cases. It is available now via Google AI Studio and the Gemini API.</p>
<p>The overwhelmingly positive feedback from developers regarding 1.5 Flash-8B has been incredible, and we will continue to shape our experimental to production release pipeline based on developer insights.</p>
<p>We're excited about these updates and look forward to seeing the innovative solutions you will create with the new Gemini models! Additionally, for Gemini Advanced users, a chat-optimized version of Gemini 1.5 Pro-002 will soon be accessible.</p>