發布日期2026/03
推論速度0 tokens/s
首字延遲時間0.00s
輸入價格$2/1M
輸出價格$6/1M
排名表現
評測分析
Grok 4.20 is a specialized model for AI agents and logical reasoning, well-suited for developers and engineers requiring complex automation. However, it is not a primary choice for general conversation or open-ended consulting.
強項
- Excellent AI agent capabilities (ranked #8)—very strong in automation and complex task execution
- Solid logical reasoning (ranked #33)—capable of handling demanding analytical, mathematical, and deductive problems
- Well-suited for automated workflows requiring autonomous decision-making, planning, and sequential execution
弱項
- Weak general conversation (ranked #51)—not ideal for open-ended chat or daily Q&A
- Suboptimal open-ended context comprehension and long-form conversation handling
- Not a general-purpose choice—requires alternative models if an all-around assistant is needed
適用情境
Building AI agents to automate complex tasks (research, analysis, coding)Solving logical reasoning, STEM, and data analysis problemsWorkflows requiring autonomous decision-making and execution—test automation, data pipelines, autonomous planning
指南與影片
Access grok.com or use the Grok app on iOS/Android, logging in with an X account. Grok 4.20 is an AI chatbot capable of accessing real-time information from the X platform, supporting both image and text processing. Use it for content writing, coding, data analysis, or real-time information retrieval. Tip: select Regular Mode for accurate outputs or Fun Mode for a sarcastic tone.