Google Releases Gemini Pro API for Developers and Enterprises
Google makes its Gemini Pro AI model available to developers through a new API and Google AI Studio. The release includes robust features like a 32K context window, multimodal vision capabilities, and support for 38 languages.
Google officially opens up its Gemini Pro AI model to developers and enterprises through the new Gemini API. This release allows creators to integrate the model into their own applications, building on the recent debut of the Gemini family which includes Ultra, Pro, and Nano sizes. A specialized version of Gemini Pro currently powers Google's Bard chatbot, and now external developers gain access to this same powerful technology.
The Gemini Pro model outperforms similarly-sized models on standard research benchmarks and offers a 32K context window for text inputs. Developers receive access to advanced features such as function calling, embeddings, semantic retrieval, and custom knowledge grounding. Additionally, a dedicated multimodal endpoint called Gemini Pro Vision is available to process both text and images together, with support spanning 38 languages across over 180 countries.
To accelerate development, Google launches Google AI Studio, a free web-based tool for prompt engineering and rapid prototyping. The platform provides a generous free tier that allows 60 requests per minute and seamlessly exports code to various integrated development environments. Furthermore, Google provides official software development kits for Python, Android (Kotlin), Node.js, Swift, and JavaScript to ensure applications run smoothly across virtually any platform.