Freemiumchat

Grok

Real-time AI assistant with frontier coding and agent capabilities

Grok is SpaceXAI's AI assistant for real-time research, coding, and agentic work. Grok 4.6 adds a 500K-token API context window, text and image input, configurable reasoning, and stronger long-running task performance.

4.4

Try Grok Free →

Opens official website

AI assistantreal-time searchX integrationcodingAI agentslong contextmultimodal

Live change feed

Recent verified changes

Shared with the homepage and tool pages

criticalgrokAug 12, 2026

Grok 4.6 launches for long-running agents, coding, and knowledge work

SpaceXAI released Grok 4.6 with a 500K-token context window, text and image input, configurable reasoning, and stronger support for long-running agent and coding tasks. Standard API pricing is $2 per million input tokens, $0.50 cached input, and $6 output below 200K prompt tokens; long-context rates are $4, $1, and $12.

Verify official source →

About Grok

Grok is SpaceXAI's general-purpose assistant and model family for research, coding, knowledge work, and real-time information workflows. Grok 4.6, released August 12, 2026, focuses on long-running agents and ambitious interactive work. The API model accepts text and images, provides text output, supports function calling, structured outputs, and configurable reasoning, and has a 500K-token context window. Standard API pricing below 200K prompt tokens is $2 per million input tokens, $0.50 cached input, and $6 output; long-context requests are billed at $4, $1, and $12. AI Linkbase treats SpaceXAI's benchmark results as evidence for re-evaluation, not as sufficient grounds for an automatic rating or ranking change.

✅ Pros

+Native real-time X and web search tools for current research workflows
+Grok 4.6 supports a 500K-token API context window with text and image input
+Standard API pricing starts at $2 input, $0.50 cached input, and $6 output per 1M tokens below 200K prompt tokens
+Configurable reasoning plus function calling and structured outputs
+Available through the SpaceXAI API, Grok Build, Cursor, OpenRouter, Vercel, and Cloudflare at launch

❌ Cons

−Requests above 200K prompt tokens use higher long-context pricing: $4 input, $1 cached input, and $12 output per 1M tokens
−No built-in access to real-time events unless Web Search or X Search tools are enabled
−Vendor-reported benchmark gains still require independent task testing before procurement or ranking changes
−Consumer Grok plan access and usage limits are separate from API availability and pricing
−Enterprise buyers should review safety, governance, and data-handling controls for agentic deployments

Pricing Plans

Consumer subscriptions and API billing are separate. Grok 4.6 API: $2 input / $0.50 cached input / $6 output per 1M tokens below 200K prompt tokens; $4 / $1 / $12 above 200K. Fast service is priced separately.

✅ Best For

Real-time web and X researchLong-running coding and agent tasksKnowledge workLarge-context API workflowsInteractive app prototyping

⚠️ Not Recommended For

Teams that need rankings based only on independently reproduced benchmarksWorkflows that require predictable low cost above 200K prompt tokensOrganizations that have not reviewed agent permissions, logging, and data controls