Модели xAI

Grok 4.5 is an xAI model for coding, agentic tasks and knowledge work. It supports configurable reasoning effort, so applications can trade response time for deeper analysis. The model accepts text and image inputs and works well for complex questions, software tasks and multi-step tool workflows.

Grok 4.7 is an xAI frontier model for coding, agentic tasks and knowledge work. It accepts text and image inputs, supports a 500,000-token context window and configurable reasoning effort, and can work through long, multi-step tasks. Its API route is distinct from earlier Grok 4 variants.

Grok 4.20 Beta is xAI’s latest flagship model, offering industry-leading speed and advanced agentic tool-calling capabilities. It features one of the lowest hallucination rates on the market and strong prompt adherence, enabling consistently accurate and reliable responses.

Grok Code Fast 1 is a speedy and economical reasoning model that excels at agentic coding. With reasoning traces visible in the response, developers can steer Grok Code for high-quality work flows.

A lightweight, high-speed model from xAI, engineered for logic-based tasks that do not require deep domain knowledge. Grok-3-mini is optimized for rapid response and efficient reasoning, making it ideal for applications where speed and concise logic are prioritized over extensive context or specialized expertise. The model exposes raw thinking traces, providing transparency into its decision-making process and enabling advanced debugging or educational use cases.

Grok-2-Aurora is an autoregressive, mixture-of-experts model from xAI, trained on billions of text and image examples. It excels at photorealistic rendering, accurately following text instructions, and complex scene generation, leveraging deep world understanding built during training.

xAI’s Grok 2 Vision 1212 is a next-generation vision-language model designed for advanced image-based AI applications. It features robust visual comprehension, refined instruction-following, and strong multilingual support. The model excels at object recognition, style analysis, and visual reasoning, empowering developers to build intuitive, visually aware applications. Enhanced steerability and reasoning capabilities make it a solid foundation for next-generation image solutions. For more details, see the official [xAI announcement](https://x.ai/blog/grok-1212).

Grok 4.20 is a reasoning model from SpaceXAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherance, delivering consistently precise and truthful responses.

Grok 4.20 Reasoning is the reasoning configuration of the Grok 4.20 family. It spends additional inference effort on multi-step questions, coding and analysis, making it suitable when a more deliberate answer is useful. It is exposed separately from the faster non-reasoning route so users can choose the desired behavior.

Grok 4.20 Instant is the non-reasoning configuration of the Grok 4.20 family. It favors quick responses for everyday conversation, drafting and straightforward questions. It uses a distinct upstream route from Grok 4.20 Reasoning, allowing applications to select speed or deeper reasoning explicitly.

Grok 4.3 is a reasoning model from SpaceXAI. It accepts text and image inputs with text output, and is suited for agentic workflows, instruction-following tasks, and applications requiring high factual accuracy. Reasoning can be configured between none/low/medium/high (default low) effort levels.

Grok 4.6 is an xAI frontier model for coding, agentic tasks and knowledge work. It accepts text and image inputs, has a 500,000-token context window and supports low through extra-high reasoning effort. It is suited to long documents, software engineering and multi-step analysis.

Grok Imagine Image 2.0 is xAI’s image generation and editing model. It creates images from text and transforms reference images with improved instruction following, typography and visual consistency. The Kie integration exposes separate generation and editing routes; editing can use multiple reference images and supports common aspect ratios.