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. Strong all-purpose performance

• Designed as Google's balanced middle-tier model.
• Handles a wide range of tasks: reasoning, writing, coding, and problem-solving.

2. Natively multimodal understanding

• Trained from the ground up on text, images, audio, and video.
• More consistent multimodal reasoning than stitched-together architectures.

3. Great cost-to-capability ratio

• Offers much of Gemini Ultra's reasoning quality at a fraction of the cost.
• Strong default choice for large-scale production workloads.

4. Reliable reasoning and factual performance

• Performs well on benchmarks like MMLU, MMMU, and code reasoning.
• Handles long-form analysis, multi-step reasoning, and structured problem solving.

5. Advanced coding capabilities

• Supports major languages such as Python, Java, C++, Go.
• Generates, edits, debugs, and explains code with high accuracy.
• Powers advanced coding systems like AlphaCode 2.

6. Efficient and scalable

• Optimized for Google TPUs for lower latency and faster inference.
• Suitable for batch workloads, agents, and complex multi-step pipelines.

7. Strong multimodal reasoning

• Understands math, physics, and scientific diagrams.
• Handles mixed data inputs (charts + text, screenshots + instructions, etc.).

8. Enterprise-ready reliability

• Available through Google AI Studio and Vertex AI.
• Benefits from enterprise-grade governance, safety, privacy, and compliance.