OpenAI has unveiled a new model, referred to as GPT-5.3-Codex-Spark, for real-time coding. The model is reported to deliver high speeds, with some outlets stating it can generate over 1,000 tokens per second, and is said to be up to 15 times faster than previous models. The model's capabilities and specifications, including its use of Cerebras chips and a 128k context, have been detailed by various outlets.