# Grok 4.7

SpaceXAI, released 2026-09-21

**Verdict: conditional.** Grok 4.7 keeps Grok 4.6's price and speed, and xAI says it works longer on hard tasks and checks its own work more carefully. Conditional because every published score is xAI's own, prompts of 200K tokens or more bill at double the headline rate, and the Batch API is not supported. Right for coding teams already working in Cursor or Grok Build. Last verified 2026-10-02.

## Pricing

- Input: $2.00 / 1M tokens
- Output: $6.00 / 1M tokens
- Context: 500K ctx; 200K+ $4/$12; fast variant 2x price

## Strengths

- Coding agents in Cursor and Grok Build, where it is available from launch day
- Low-cost frontier-class calls at $2 input and $6 output per 1M tokens under 200K, cached input $0.50
- Long-running agent tasks: xAI says it works longer on difficult tasks and verifies its own work more carefully
- Tuning depth per call with reasoning effort low, medium, high or xhigh
- Prompts of 200K tokens or more budgeted off the headline rate: input, cache and output all double to $4, $1 and $12
- Batch workloads: the Batch API is not supported
- Decisions that need independent benchmarks: every published score is xAI's own

## Benchmarks

- DeepSWE v1.1, high effort (xAI-reported): 71.0% (https://x.ai/news/grok-4-7)
- CursorBench 4.0 (xAI-reported): 46.3% (https://x.ai/news/grok-4-7)
- Terminal-Bench 4.0 (xAI-reported): 37.6% (https://x.ai/news/grok-4-7)
- GDPval, xhigh effort (xAI-reported): 1695 (https://x.ai/news/grok-4-7)

## Sources

- [xAI: Grok 4.7 (Sep 21, 2026)](https://x.ai/news/grok-4-7)
- [xAI docs: Grok 4.7 model (also Models and pricing)](https://docs.x.ai/docs/models/grok-4.7)
- [xAI models and pricing](https://docs.x.ai/docs/models)

Canonical: https://neomanex.com/models/grok-4-7