Grok gives developers API access to text, image, video and voice models. The Models and Pricing pages list each model's context window and per-token prices. The flagship is grok-4.7, which the page describes as its model for code and everything else, with agentic tool calling, configurable reasoning and a 500k token context window.
The platform covers several kinds of work. Text models handle chat, reasoning and structured outputs, and accept image input. Imagine models generate and edit images and videos. Voice models provide real-time speech-to-speech, text-to-speech and speech-to-text. Models can also call server-side tools such as web search, X search and code execution, which are billed separately.
The documentation is written for developers who build with an API. It explains model aliases, so users can pin a specific release or follow the latest one. It also notes that models have no knowledge of current events beyond their training data unless search tools are enabled.
Pricing is usage-based and set in US dollars. grok-4.7 costs $2.00 per million input tokens and $6.00 per million output tokens for short contexts, with higher rates once a request reaches 200k tokens. Speech-to-speech is $0.08 per minute, text-to-speech is $15.00 per million characters, and image generation starts at $0.02 per image. Requests sent to the US regional endpoint are billed at a 10 percent premium.
