1. Strongest Agentic Coding Model

• State-of-the-art on Terminal-Bench 2.0 (82.7%), Expert-SWE (73.1%), and SWE-Bench Pro (58.6%), outperforming GPT-5.4 on complex coding tasks.
• Holds context across large systems, reasons through ambiguous failures, and carries changes through surrounding codebases with fewer tokens.

2. Higher Intelligence at GPT-5.4 Latency

• Co-designed, trained, and served on NVIDIA GB200/GB300 NVL72 systems to match GPT-5.4 per-token latency while performing at a significantly higher level.
• Uses fewer tokens to complete the same tasks, making it more efficient as well as more capable.

3. Powerful for Knowledge Work & Computer Use

• Scores 84.9% on GDPval (44 occupations) and 78.7% on OSWorld-Verified for autonomous computer operation.
• Excels at generating documents, spreadsheets, and reports; naturally moves across finding information, using tools, and checking output.

4. Scientific Research Co-Scientist

• Leading performance on GeneBench, BixBench, and FrontierMath; helped discover a new proof about Ramsey numbers verified in Lean.
• Strong enough to meaningfully accelerate progress at the frontiers of biomedical and mathematical research.

1. High reasoning performance

• Retains strong reasoning capabilities despite being a smaller, faster model.
• Suitable for tasks requiring accurate logic and structured thinking.

2. Fast and cost-efficient

• Optimized for speed, making it ideal for real-time or high-volume workloads.
• Far cheaper than GPT-5 while maintaining solid capability.

3. Great for well-defined tasks

• Excels when prompts are precise and objectives are clearly specified.
• More predictable and stable for deterministic workflows.

4. Multimodal input

• Accepts text + image as input.
• Outputs text only.

5. Tool support

• Works with Web Search, File Search, Code Interpreter, MCP.
• (Does not support Image Generation as a tool and does not support Computer Use.)