Google has released Gemini Flash 2.0, the latest version of its large language model (LLM) for its Gemini AI assistant. Flash 2.0 was previously a beta, but is now available to all users via a web app and Android app.

Flash 2.0 is significantly faster than its predecessor. Gemini says Flash 2.0 is the model for everyday tasks. Highlights of Flash 2.0 include increased performance, faster time to first token (TTFT), multi-modal understanding, code rewriting capabilities, the ability to execute complex instructions, and function calls.

Paid Gemini Advanced users will still get access to 1 million contextual tokens, upload up to 1,500 pages, and early access to features like Deep Research and Gems.

Flash 2.0 is the default model for new conversations, but Google has confirmed that 1.5 Flash and 1.5 Pro will continue to be used for existing conversations for a few more weeks.