OpenAI: GPT-5.1-Codex-Max

GPT-5.1-Codex-Max, OpenAI’s latest agentic coding model, demonstrates exceptional reliability with a 100% success rate across all benchmarks, indicating consistent operational stability. Its speed performance is moderate, ranking in the 25th percentile, while its pricing is positioned at premium levels, falling into the 6th percentile. The model exhibits strong performance in several key areas. It achieved perfect accuracy in General Knowledge and Ethics, with the latter also being the most accurate and fastest among models at its price point. Instruction Following is a significant strength, with 85% accuracy, placing it in the 95th percentile. Coding performance is also robust at 95% accuracy (93rd percentile), and Mathematics shows solid results at 93% accuracy (73rd percentile). Reasoning capabilities are fair at 80% accuracy. A notable weakness is its performance in Hallucinations, where it achieved only 2.0% accuracy, indicating a tendency to provide definitive answers to fictional concepts rather than acknowledging uncertainty. Email Classification also shows room for improvement at 97% accuracy, placing it in the 42nd percentile. Overall, GPT-5.1-Codex-Max excels in complex reasoning and knowledge-based tasks, particularly within its intended domain of software development, but users should be mindful of its hallucination tendencies.