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. Flagship-level reasoning and math performance

• Designed for world-class reasoning depth, precision, and multi-step logical chains.
• Excels at STEM, mathematics, symbolic operations, proofs, and analytical workloads.

2. Powerful multimodal understanding

• Supports text, images, and other modalities.
• Handles cross-modal reasoning tasks requiring context synthesis.

3. Extreme capability across diverse tasks

• Positioned as a top-tier 'jack of all trades' model.
• Strong in natural language, coding, knowledge retrieval, and structured generation.

4. Large 256K context window

• Enables analysis of long documents, entire codebases, multi-document packs, and extensive agent sessions.
• Supports workloads that require persistent reasoning across large inputs.

5. Advanced developer tooling support

• Function calling for tool-augmented workflows.
• Structured outputs for predictable, schema-controlled generation.
• Integrates smoothly with agents and complex automation pipelines.

6. Efficient caching for cost reduction

• Cached input tokens discounted to $0.75 / 1M tokens.
• Encourages RAG, retrieval pipelines, and multi-step conversational workflows.

7. Production-ready performance

• Stable rate limits: 480 requests per minute.
• High token throughput: 2,000,000 tokens per minute.
• Available across multiple xAI regional clusters.

8. Optional Live Search augmentation

• Add-on: $25 per 1K sources.
• Enhances factual accuracy and real-time information retrieval.