X @外汇交易员

Model Performance - OpenAI's GPT OSS 120B achieves a speed of approximately 43-45 tokens/second (t/s) on an Apple M3 Ultra 512G [1] - GPT-4.5 provides a favorable evaluation of GPT OSS's responses, indicating strong performance [1] Product Release & Capabilities - OpenAI releases gpt-oss, an open model designed to perform at the level of o4-mini [1] - The open model can run on a high-end laptop, with a smaller version capable of running on a phone [1]