xAIProprietaryReasoningCoding
Grok
xAI’s Grok 4.6 — frontier coding and long-running agentic model (API id grok-4.6) with 500K context, vision, and strong tool use. Available via the xAI API, Grok Build, Cursor, and partners such as OpenRouter.
Vision · Tool calling · Thinking · Coding
Last reviewed: 22 August 2026
Overview
xAI’s Grok 4.6 — frontier coding and long-running agentic model (API id grok-4.6) with 500K context, vision, and strong tool use. Available via the xAI API, Grok Build, Cursor, and partners such as OpenRouter.
Capabilities
- Vision: Yes
- Audio: No
- Tool calling: Yes
- Thinking: Yes
- MCP: No
- Coding: Yes
- Structured output: Yes
Technical specifications
- Provider
- xAI
- License
- Proprietary (select older open releases exist separately)
- Context window
- 500K
- Architecture
- Transformer (proprietary)
- Release
- 2026-08
- Modalities
- Text, Image
- Vision
- Yes
- Audio
- No
- Tool calling
- Yes
- Thinking
- Yes
- MCP
- No
- Open weights
- No
- API
- Yes
- Pricing (input)
- ~$2 per 1M tokens; fast variant 2×
- Pricing (output)
- ~$6 per 1M tokens; fast variant 2×
Supported modalities
Text · Image
Context window
500K (500,000 tokens)
Pricing
Input: ~$2 per 1M tokens; fast variant 2×
Output: ~$6 per 1M tokens; fast variant 2×
Official Grok 4.6 post lists $2/$6 starting rates and a 2× fast variant. Cached input $0.50/M (up from $0.30 on 4.5). Rates double above 200K prompt tokens. Verify docs.x.ai for live tiers.
Availability
API: Yes
Chat UI: Yes
Open weights: No
API model grok-4.6; also in Cursor, Grok Build, and GitHub Copilot (Business/Enterprise policy off by default). First-week 2× included usage in Cursor/Grok Build from 2026-08-12.
Use cases
- Long-running agentic coding and knowledge work
- Interactive / visual application first passes
- Tool-heavy software engineering agents
- Vision + text workflows via xAI API
Strengths
- Tuned for multi-step agents vs Grok 4.5
- 500K context for large codebases and sessions
- Same headline $2/$6 API price as 4.5; Cursor distribution
Limitations
- Closed weights
- Ecosystem smaller than OpenAI/Anthropic
- Long-prompt tier doubles token rates above 200K
Related guides
Related benchmarks
Related research
- Introducing Grok 4.6
Original Paper
Related GitHub
Related tools
Related rankings
Related comparisons
Companies
Explore more models
- Claude FableAnthropicAnthropic’s Claude Fable 5.1 — the most capable widely released Claude for long-horizon agents, deep reasoning, and demanding coding workflows. Mythos 5.1 is the same underlying model with relaxed cyber/life-sciences safeguards for trusted-access programs.
- Claude OpusAnthropicAnthropic’s Claude Opus 5 tier for complex agentic coding, enterprise work, long-context analysis, and careful instruction following. Claude Fable 5.1 sits above Opus for peak widely released capability.
- Claude SonnetAnthropicAnthropic’s Claude Sonnet 5 tier — best combination of speed and intelligence for most production agents and coding, at lower cost than Opus.
- Gemini 3.1 ProGoogleGoogle’s current Pro-class Gemini for hard reasoning and native multimodal work. Prefer API id gemini-3.1-pro-preview; Gemini 3.5 Pro remains partner-testing. Legacy gemini-2.5-pro is scheduled for shutdown Oct 16, 2026.
- GPT-5.6OpenAIOpenAI’s GPT-5.6 family (Sol flagship, Terra balanced, Luna cost-efficient) for complex reasoning, coding, multimodal understanding, and agentic tool use. The gpt-5.6 API alias routes to Sol.
- GPT-6 AstraOpenAIOpenAI’s GPT-6 Astra peak model (API id gpt-6-astra) for computer use, coding agents, professional artifacts, science, and cybersecurity-adjacent defender workflows. Staged launch 2026-09-03; GPT-5.6 Sol/Terra/Luna remain the volume and cost-routing family.