GPT-4o mini OpenAI's most cost-effective small model, outperforming GPT-3.5 Turbo and other small models. Supports multiple languages and features a 128K token context window.