On February 13, OpenAI introduced the GPT-5.3-Codex-Spark research preview, a streamlined version of GPT-5.3-Codex. According to BlockBeats, this marks OpenAI's first model specifically designed for real-time coding, enabling targeted editing, logic restructuring, and interface optimization with immediate result viewing. Codex-Spark is optimized for ultra-low latency hardware, delivering nearly instantaneous responses, capable of processing over 1,000 tokens per second while effectively handling practical coding tasks.