The performance of Jalapeño was evaluated using the InferenceX benchmarking platform, showing significant improvements in AI task completion per watt and reduced end-to-end latency compared to existing systems. OpenAI plans to continue developing this chip alongside its existing partnerships.