#GPT-5.6

OpenAI's GPT-5.6 Sol Ultra model autonomously proved the cycle double cover conjecture, a 50-year-old open problem in graph theory, in under one hour using 64 sub-agents. The model generated a 3-page PDF proof and OpenAI released the exact prompt used. This demonstrates that large language models can now tackle long-standing unsolved mathematical problems through advanced multi-agent reasoning, potentially transforming how mathematical research is conducted. It also validates the effectiveness of carefully crafted prompts and sub-agent architectures for complex reasoning. The model used 64 sub-agents working in parallel, transforming the problem into one about edge labelings and linear equations over finite fields. The prompt explicitly defined acceptance criteria, definitions, boundary conditions, and failure cases, while requiring dynamic sub-agent allocation and independent verification.

OpenAI has launched the GPT-5.6 series, featuring three models: Sol (flagship), Terra (balanced), and Luna (cost-efficient). The flagship Sol model achieves a state-of-the-art score of 7.8% on the ARC-AGI-3 benchmark, becoming the first verified frontier model to beat an ARC-AGI-3 game. This release marks a significant leap in AI reasoning and agentic capabilities, particularly on the challenging ARC-AGI-3 benchmark that measures interactive reasoning and adaptation. The GPT-5.6 series strengthens OpenAI's competitive position against rivals like Anthropic's Fable 5, and the introduction of max/ultra reasoning, multi-agent collaboration, and Programmatic Tool Calling enables more efficient complex task execution. The GPT-5.6 series includes Sol, Terra, and Luna models, with Sol being the most capable. GPT-5.6 Sol sets a new SOTA on ARC-AGI-3 at 7.8% using max reasoning effort, and also improves performance on coding, science, and cybersecurity tasks. The developer guide highlights improved intent understanding and image detail preservation, while GPT-5.4 is scheduled to be deprecated on July 23.