Gemini, Google’s latest flagship GenAI model family, is now available to Google Cloud customers using Vertex AI, following its recent launch for Bard and the Pixel 8 Pro. The Gemini Pro, a more lightweight version of the Gemini Ultra model, is now accessible in public preview in Vertex AI through the new Gemini Pro API. The API is currently free to use within certain limits and supports 38 languages and regions, along with features like chat functionality and filtering.
“Gemini is a cutting-edge natively multimodal model with advanced reasoning and coding abilities,” stated Google Cloud CEO Thomas Kurian during a press briefing on Tuesday. “Developers will now be able to build their own applications using it.”
Gemini Pro API
The Gemini Pro API in Vertex accepts text as input and generates text as output by default, similar to other generative text model APIs. Additionally, a new endpoint called Gemini Pro Vision, also launching today in preview, can process both text and imagery, including photos and video, and produce text-based outputs.
Users can now harness the model’s image comprehension and processing abilities within Vertex AI. They also have the option to customize Gemini Pro to specific contexts and use cases using the same fine-tuning tools available for other models hosted on Vertex.
Citation checking, another existing Vertex AI capability, now also supports Gemini Pro, providing an additional fact-checking measure by highlighting the sources of information used by the model.
Input for Gemini Pro on Vertex AI is priced at $0.0025 per character, while output is priced at $0.00005 per character. For a limited time, Gemini Pro is free to try for Vertex AI customers until early next year.
Enhancements to Vertex AI
Google is introducing several new features to Vertex AI to compete with other platforms like Bedrock. These features include the ability to use Gemini Pro to power custom-built conversational voice and chat agents, as well as driving search summarization, recommendation, and answer generation features within Vertex AI.
Additional features include Automatic Side by Side (Auto SxS) to evaluate models in an automated fashion and the addition of models from third parties such as Mistral and Meta. Google is also extending its indemnification policy to include outputs from PaLM 2 and its Imagen models to legally defend eligible customers implicated in lawsuits over IP disputes involving those models’ outputs.
While Google’s Vertex AI indemnification policy currently does not cover customers using the Gemini Pro API, the company plans to extend it once the Gemini Pro API launches publicly.