Google DeepMind Releases Upgraded Gemini 2.0 Flash, Pro, and Flash-Lite Models
Google DeepMind makes the updated Gemini 2.0 Flash generally available for developers while introducing new Pro and Flash-Lite models. These updates expand multimodal capabilities and offer enhanced performance for coding and complex tasks.
Google DeepMind makes the updated Gemini 2.0 Flash model generally available to developers through the Gemini API in Google AI Studio and Vertex AI. This highly efficient workhorse model supports high-volume tasks and features a massive context window of one million tokens. The updated Flash model already powers the Gemini app on desktop and mobile, and it soon promises to deliver image generation and text-to-speech capabilities.
The company also introduces an experimental version of Gemini 2.0 Pro, which stands as Google's best model yet for coding performance and handling complex prompts. Advanced users access this powerful model directly within the Gemini app, while developers can integrate it via Google AI Studio and Vertex AI. Additionally, a brand new Gemini 2.0 Flash-Lite model enters public preview to serve as the most cost-efficient option in the growing Gemini 2.0 family.
All of these newly released models currently feature multimodal input with text output, and Google plans to add more modalities in the coming months. The 2.0 Flash Thinking Experimental model also becomes accessible to regular Gemini app users through the model dropdown menu on both desktop and mobile platforms. Google DeepMind continues to work on further updates and improved capabilities for the entire Gemini 2.0 lineup as the agentic era progresses.