Google plans to launch a large-scale machine learning model codenamed Gemini, equipped with multimodal capabilities, suitable for tasks such as chart analysis, text or voice commands, and more. Gemini is a collaborative effort between Google and DeepMind, with a team of 26 directors overseeing different groups. The model is set to debut at a fall launch event, aiming to compete with OpenAI's GPT-4 and enhance Google's competitiveness in the field of artificial intelligence.