Google Releases Gemini 2.0 Flash — Cheapest Frontier Multimodal Model

Google launched Gemini 2.0 Flash, their fastest and cheapest advanced AI model yet. Despite being built for speed and cost efficiency, it punches above its weight — it outperforms OpenAI's GPT-4o-mini while costing less. It can process up to one million tokens at once, which is roughly the length of ten novels, letting it handle enormous documents or long codebases in a single request. The model understands text, images, audio, and video, and can run code directly. It was specifically designed to be the backbone of AI agents — software that takes actions, browses the web, and completes multi-step tasks automatically. At just $0.075 per million tokens it became the most affordable frontier-class model available, democratizing access to capable AI for startups and developers working at scale.