"GPT-5.5 represents a step toward AI systems that can complete complex, multi-step tasks on a computer without human guidance, showcasing significant improvements in autonomous capabilities."
"On SWE-Bench Pro, GPT-5.5 resolves 58.6% of tasks end-to-end in a single pass, demonstrating its effectiveness in real-world GitHub issue resolution."
"Developers reported that GPT-5.5 has a better understanding of the 'shape' of a software system, allowing it to identify failures and necessary fixes more effectively."
OpenAI launched GPT-5.5, a powerful AI system that enhances the Codex coding agent's ability to perform complex digital tasks. It excels in scientific work, generating and testing hypotheses. The system shows significant improvements in autonomous capabilities, completing multi-step tasks without human guidance. GPT-5.5 outperforms previous models on benchmarks like Terminal-Bench 2.0 and OSWorld-Verified, indicating its superior command-line workflow and independent operation. With a growing user base, it enables Codex to produce high-quality code and tackle projects with senior software engineer-level judgment.
Read at Fast Company
Unable to calculate read time
Collection
[
|
...
]