Gemini 1.5 Pro updates, 1.5 Flash debut and 2 new Gemma models

Seize an API key in Google AI Studio, and get began with the Gemini API Cookbook.

Final December, we released Gemini 1.0 and made the Gemini API available in Google AI Studio and Vertex AI for enterprise clients. In February, we released Gemini 1.5 Pro, the primary mannequin with a 1 million token context window. Final month, we added capabilities for native audio understanding, system directions, JSON mode and extra.

In parallel, we launched Gemma, our household of light-weight open fashions constructed from the identical analysis and expertise used to create the Gemini fashions. Since February, the 2B and 7B fashions have been downloaded thousands and thousands of occasions, and we added variants final month for builders and researchers, like CodeGemma and RecurrentGemma.

Now Gemini and Gemma are getting even higher. Right this moment, we’re updating Gemini 1.5 Professional, introducing 1.5 Flash, rolling out new developer options and including two new Gemma fashions.

Listed here are particulars on all of the updates.

Gemini 1.5 Professional enhancements and a brand new 1.5 Flash mannequin

• Gemini 1.5 Professional: We made a collection of high quality enhancements throughout key use instances, equivalent to translation, coding, reasoning and extra. You’ll see these updates within the mannequin beginning right now, which ought to enable you sort out even broader and extra complicated duties.
• Gemini 1.5 Flash: This smaller Gemini mannequin is optimized for narrower or high-frequency duties the place the pace of the mannequin’s response time issues essentially the most.
• Availability: Each fashions can be found right now in additional than 200 countries and territories in preview and might be typically out there in June.
• Natively multimodal with lengthy context: Each 1.5 Professional and 1.5 Flash include our 1 million token context window and let you interleave textual content, pictures, audio and video as inputs. To get entry to 1.5 Professional with a 2 million token context window, be part of the waitlist in Google AI Studio or in Vertex AI for Google Cloud clients.