To completely no person’s shock, AI took middle stage at Google I/O, the corporate’s annual builders convention. And we acquired not one, however two new Gemini AI fashions, 1.5 Flash and Nano, together with some extra details about the 1.5 Professional mannequin introduced in February.
The excellent news is each Gemini 1.5 Professional and 1.5 Flash can be found in preview globally now. However if you wish to check out Google’s new 2 million token context window, you will have to affix the waitlist right here.
Extra from Google I/O 2024
Gemini 1.5 Professional and Flash are each natively multimodal, which suggests you may immediate them with textual content, photographs and movies. Google additionally introduced new updates to the Gemini 1.5 API, together with video body extraction, parallel operate calling and context caching — which, amongst different issues, means you will solely must add your recordsdata to the fashions as soon as.
Gemini 1.5 Professional: Probably the most superior
Gemini 1.5 Professional was first launched for early testing in February, the place it had a most 128K token window. Now, it is out there extra extensively for builders and has a baseline token window of 1 million, with an choice to enroll in a 2 million token window.
- Price: $7 per 1 million tokens; $3.50 per 1 million tokens for prompts as much as 128K
- Finest for: Complicated duties, producing higher-quality outcomes
- Response time: Slower than 1.5 Flash
- Token window: 1 million out there now, 2 million out there by way of waitlist
Gemini 1.5 Flash: Constructed for pace
Gemini 1.5 Flash is a brand new, leaner mannequin launched at Google I/O. Its most important promoting level is its response time, which is notably sooner than 1.5 Professional. Gemini 1.5 Flash is out there now globally. This is what you must learn about it.
- Price: Beginning at $0.35 per million tokens
- Finest for: Slim and high-frequency duties you want accomplished ASAP
- Response time: Sooner than 1.5 Professional
- Token window: 1 million out there now, 2 million out there by way of waitlist
Gemini Nano: Coming quickly to a Pixel telephone close to you
Gemini Nano with Multimodality goes to be coming to Google Pixel smartphones later this yr. The cellular AI mannequin will have the ability to course of textual content, photographs, audio and speech. Android was the primary cellular working system to incorporate a built-in, on-device AI mannequin, so this subsequent replace within the new Pixel will increase its AI capabilities.
Through the I/O keynote, Dave Burke, Vice President of Engineering for Android, talked about how Nano on Pixels goes to be bringing some extra firepower to Android’s accessibility choices. Particularly, Nano will likely be a part of updates to Android’s TalkBack characteristic, which gives auditory and textual content descriptions of photographs missing alt textual content. He additionally showcased how a brand new Nano characteristic will assist name out suspicious or spam telephone calls.
Editor’s observe: CNET is utilizing an AI engine to assist create a handful of tales. Evaluations of AI merchandise like this, similar to CNET’s different hands-on opinions, are written by our human staff of in-house specialists. For extra, see CNET’s AI coverage and how we check AI.
