GPT-5.5 (high) is a flagship large language model developed by OpenAI and released on April 23, 2026. As a fully retrained base model, it represents a significant architectural shift from the incremental GPT-5.1 through GPT-5.4 updates. The "high" designation refers to a specific reasoning effort tier that utilizes parallel test-time compute to achieve higher intelligence scores on complex tasks while maintaining the per-token latency of earlier iterations.
Designed as a "new class of intelligence for real work," the model is natively multimodal, supporting the processing and generation of text, images, audio, and video within a single system. It introduces advanced agentic capabilities, particularly in the domains of computer use—the ability to autonomously operate desktop environments and browsers—and agentic coding. Benchmark results indicate a success rate of 82.7% on Terminal-Bench 2.0 and expert-level performance across a wide range of professional knowledge domains.
The model features a significantly expanded context window of 922,000 tokens, enabling it to process and maintain context across long-horizon research and multi-step workflows. OpenAI has also emphasized token efficiency, noting that GPT-5.5 often requires fewer tokens than its predecessors to complete the same tasks. This efficiency is paired with a revised pricing structure designed for enterprise-scale productivity and complex autonomous agents.