Edgar Cervantes / Android Authority
TL;DR
- Google is making Gemini 2.0 Flash usually obtainable by means of the Gemini API in AI Studio and Vertex AI.
- Gemini app customers will now be capable of entry 2.0 Flash Considering Experimental from the mannequin dropdown on desktop and cell.
- The corporate can also be rolling out Gemini 2.0 Professional and Flash-Lite.
Google first launched Gemini 2.0 Flash as an experiment again in December. A month later, 2.0 Flash had its experimental standing dropped and the corporate improved its 2.0 Flash Considering Experimental mannequin in Google AI Studio. Google is now making these fashions extra available whereas rolling out just a few further AI fashions.
As we speak, the Mountain View-based agency introduced that Gemini 2.0 Flash — its workhorse mannequin for builders — is being made usually obtainable by means of the Gemini API in Google AI Studio and Vertex AI. This comes solely every week after it had made the up to date mannequin obtainable to all Gemini app customers on desktop and cell.
On prime of this, Gemini app customers will now be capable of entry Gemini 2.0 Flash Considering Experimental from the mannequin dropdown on desktop and cell. For these unfamiliar with this model of Gemini, it’s an enhanced reasoning mannequin that may present its thought course of and cause by means of extra complicated issues. You possibly can take a look at an in depth efficiency comparability in Google’s weblog.
If you happen to’re not overwhelmed by all of those totally different Gemini fashions but, then excellent news, Google is releasing much more. Beginning as we speak, the tech large is launching an experimental model of Gemini 2.0 Professional. Google claims that this mannequin “has the strongest coding efficiency and talent to deal with complicated prompts, with higher understanding and reasoning of world data, than any mannequin we’ve launched up to now.” It might additionally name on instruments like Google Search and code execution.
Along with 2.0 Professional, Google can also be debuting Gemini 2.0 Flash-Lite. As its identify suggests, this mannequin is designed to be cost-efficient. Regardless of that, the agency claims that the mannequin outperforms 1.5 Flash on most benchmarks whereas being simply as quick and costing the identical quantity. If you wish to do that mannequin, it’s obtainable in public preview in Google AI Studio and Vertex AI.
Talking of value, Google presents a breakdown of the value of Gemini 2.0 Flash and a couple of.0 Flash-Lite towards 1.5 Flash within the picture above. The corporate claims the value of two.0 Flash and Flash-Lite “will be decrease than Gemini 1.5 Flash with mixed-context workloads.”
For now, these fashions solely provide multimodal (textual content, picture, and so forth) enter with textual content output. Nonetheless, Google says extra modalities can be added for basic availability within the coming months.
