Grok 4.7 Launches With $2 Input Pricing and a 500K Context Window
xAI says the new model is built for longer coding and knowledge-work tasks, with stronger verification and a new safeguard stack.
Grok 4.7 is now available from xAI, with the company positioning it as its most capable model for coding and knowledge work. The model uses a larger base model than Grok 4.6 and was trained with a longer reinforcement-learning run focused on tasks that can take hours to complete.
According to xAI's Grok 4.7 announcement, the model is designed to work longer on difficult tasks, check its own work more carefully, manage longer context, and understand the Grok Bot harness natively. It is available through Grok Build, the Grok API, Cursor, third-party coding harnesses, model routers, and cloud platforms.
Grok 4.7 Focuses on Longer Coding and Knowledge Work
xAI says Grok 4.7 improves on Grok 4.6 across several tasks. On CursorBench 4.0, the company reports a 46.3% score for Grok 4.7 at xhigh reasoning effort, compared with 40.4% for Grok 4.6 at high effort. On Terminal-Bench 4.0, xAI reports 38.0% for Grok 4.7 versus 20.3% for Grok 4.6.
The company also reports improvements in professional knowledge work. Grok 4.7 scored 1,657 on AA Briefcase v1.1, compared with 1,546 for Grok 4.6, while its EEBench electrical-engineering score rose from 53.0% to 64.0%. These figures are xAI's published benchmark results, so they should be read as vendor-reported measurements.
Grok 4.7 Keeps the $2/$6 API Pricing
Grok 4.7 starts at $2 per million input tokens and $6 per million output tokens, the same standard token pricing listed for Grok 4.6 in xAI's announcement. xAI also offers a faster Grok 4.7 variant at twice the standard token rates, but that fast version is limited to Cursor and Grok Build rather than the public xAI API.
The official Grok 4.7 API documentation lists a 500,000-token context window, text and image input, text output, function calling, web search, X search, and code execution. The API supports low, medium, high, and xhigh reasoning effort, with high as the default.
Related
- Grok Build Mode Lets SuperGrok Heavy Subscribers Build and Host Apps From a Chat Prompt
- xAI Launches Grok Imagine Image 2 With Precise Editing Tools and Arena's Second-Ranked Model
- xAI Says Grok Voice Transcribe 2.0 Doubles Accuracy at Same Price
- Grok Bot Lands on iPad and Android as the Price Drops From $300 to $20 a Month
- Grok Bot Expands Access Beyond Its Original Beta Plans
- Grok 4.6 Arrives on Google Cloud Vertex AI
- Grok 4.6 Amazon Bedrock Support Adds 500K Context
- xAI Launches Grok 4.6 at $2 per Million Tokens - Half the Price of Rival Frontier Models
New Safety and Cybersecurity Safeguards
xAI says Grok 4.7 uses an entirely new safeguard stack and was tested for refusal and jailbreak resistance. The company reports that the model allowed 3.3% of risky dual-use prompts through on its HackerBench v0.3 benchmark while rarely blocking legitimate security work. xAI has also begun giving selected cybersecurity partners invite-only access to Grok 4.7's red-team capabilities for defense research.
The launch builds on xAI's recent Grok releases, including Grok 4.6's long-running agent focus and the expansion of Grok Build to SuperGrok and X Premium plans.
Grok 4.7 is available now through the listed xAI distribution channels. The standard API pricing starts at $2 per million input tokens and $6 per million output tokens, while the model's 500K context window gives developers more room for long-running coding and knowledge-work sessions.