After announcing its Gemini family of models last week and bringing it to its Bard chatbot experience, Google is now bringing Gemini to developers by launching a series of new and updated services today. One of these services is ai Studio, formerly known as MakerSuite.
ai Studio is a web-based tool for developers that acts a bit like a gateway to the broader Gemini ecosystem, starting with Gemini Pro and then, sometime next year, Gemini Ultra as well. Using the service, developers can quickly develop Gemini-based messaging and chatbots, and then obtain API keys to use in their applications or gain access to the code to work on in a more feature-rich IDE.
It's important to note that there is a relatively generous free quota, with up to 60 requests per second, which should be enough to quickly iterate ideas without facing onerous restrictions and perhaps even enough to power some less-used applications in productions.
There's a price to pay here, though: for developers using the free tier (and that's pretty much everyone for now, as Google only plans to release a paid version early next year, likely coinciding with the GA launch of the Gemini Ultra model). , Google reviewers can see the input and output of the API and web application to “improve product quality.” However, Google notes that this data is not identified from the user's Google account and API key.
Compared to the previous version of Makersuite/ai Studio, this updated edition feels a bit more substantial. Among other things, it will offer support for both the Gemini Pro model and the Gemini Pro Vision model, allowing developers to work with both text and images.
“We've really designed it to be the fastest way to build with Gemini,” Josh Woodward, vice president of Google Labs, told me. “We really want to invite developers to play with it. It's the first version and we have a lot of tweaks we're already making now for future updates as well, but we're trying to design it in a way where people can come in and start actually building. he.”
In the web interface, developers can choose their models, adjust the temperature to control the creative range of the output, and provide examples to provide tone and style instructions. You can also adjust the model's security settings, including turning them all off.
There are also different workflows for creating structured and free-form chat messages.
Woodward noted that the team tried to design ai Studio so that even the free tier didn't feel like a trial or closed product. In fact, assuming the free tier's fee caps are sufficient for their use cases, developers can start publishing their ai Studio apps or using them through Google's API or SDKs right away.
Jeanine Banks, vice president and general manager of Google's Developer ai/”>Vertex aiGoogle's enterprise-ready generative ai development platform.
“(We have) this idea of 'growing with Google,' where you can come in, build something, run it, deploy it, let people use it, and have that generous free tier. But we're also releasing a full set of SDKs that allow developers to run and build apps with Gemini Pro that can run virtually everywhere, from the back end with support for Node.js and Python, to mobile devices, with support for Java, Kotlin and Swift, and the web, of course, with JavaScript,” he explained.
He noted that the team wants to make this transition between ai Studio and Vertex as seamless as possible. Woodward added that the SDK's strong support came directly from user feedback. “The first version we showed people said, 'I love how easy it is to give directions. Now I want to move on to the code.' And there was some kind of cliff that we had to fill in,” she told me.
Speaking of the overall ecosystem, Banks also explained that Google plans to bring Gemini to Chrome Dev Tools and Google's Firebase mobile development platform early next year.
Given the speed at which generative ai is developing, it's difficult to even predict what developers will want to use these tools for next, but Banks and Woodward emphasized that Google plans to build ai Studio as an easy on-ramp for developers of all backgrounds. levels.
“I hope that ai Studio, in some way, is not just seen as an advisory tool or something that only developers turn to, but actually, in some way, is a tool for developers. and creativity tool, where people can come up with ideas for working with these models and all the capabilities that will emerge over the next year,” Woodward said.
