Following last week’s consumer-facing announcement, Google is launching Gemini Pro for developers and enterprises through AI Studio and Vertex AI.
This first version of Gemini Pro has a 32K context window — Google plans to increase that in the future — and supports text input/output. There’s also a Gemini Pro Vision endpoint today that accepts text and image input to output text. Google also highlights how:
- “It comes with a range of features: function calling, embeddings, semantic retrieval and custom knowledge grounding and chat functionality.”
- “SDKs are available for Gemini Pro to help you build apps that run anywhere. Python, Android (Kotlin), Node.js, Swift and JavaScript are all supported.”
- “It supports 38 languages across 180+ countries and territories worldwide.”
Gemini Pro and Gemini Pro Vision are free to use today through the Google AI Studio (previously MakerSuite). Aimed at individual developers and teams, it’s a web-based tool that “enables you to quickly develop prompts and then get an API key to use in your app development.” You sign-in with your Google Account and can quickly important files from Google Drive.
The “Get code” button lets you transfer work to your IDE in cURL, JavaScript, Python, Android (Kotlin), and Swift.
A “free quota” at launch today allows 60 requests per minute, which Google says is “20x more than other free offerings.” Google will be using that free input/output to help improve the model, with data de-identified from your Google Account and API key. It “may be accessible to trained reviewers.”
Google Cloud’s Vertex AI is framed as the upgrade offering with more control and integrations. As a “fully-managed AI platform,” you can use your own data and build Gemini-powered search/conversational agents. Additionally, Google won’t train “models on inputs or outputs from Google Cloud customers.
Gemini access in Vertex AI will be free until early next year when it enters general availability.
At that time, pricing for Google AI Studio and Vertex AI will be $0.00025 per 1K characters or $0.0025 per image of input, while output costs $0.0005 per 1K characters. Google credits its efficient TPUs as making possible those prices.
Going forward, Google plans to bring Gemini to its other developer tools, including Android Studio, Colab, Firebase, and Flutter. On the consumer side, the company teased Gemini coming to Duet AI for Workspace in early 2024.
Google today also announced Imagen 2 with “significantly improved image quality,” and features like:
- “Generate high-quality, photorealistic, high-resolution, aesthetically pleasing images from natural language prompts”
- “Text rendering in multiple languages to create images with accurate text overlays”
- “Logo generation to create company or product logos and overlay them in images”
- “Visual question and answering for generating captions from images, and for getting informative text responses to questions about image details”
FTC: We use income earning auto affiliate links. More.
ncG1vNJzZmhxpKSCqLvOoKOeZpOkunB%2Bj2tqaGliZH50e8aepKKmmWK0sLvGpZxmmZliwLXBw6KmaA%3D%3D