High-speed GPT-4 variant built for large-context text workloads with enhanced tool integration. It is tuned for applications that require both strong reasoning and very high throughput, such as large copilots, code assistants, and multi-user SaaS products. The model supports long prompts and documents, which is valuable for summarization and retrieval-augmented flows. It strikes a balance between GPT-4-level quality and production-friendly performance.
