GPT-4 Turbo Preview

GPT-4 Turbo Preview is a legacy preview variant of OpenAI's GPT-4 Turbo large language model, offering a 128K context window, 4,096 max output tokens, and training data through December 2023. It supports structured JSON output, reproducible outputs, and parallel function calling, and is available via the OpenAI API and third-party inference providers such as OpenRouter. Source: supabase.