Grok 4.7 launches at Grok 4.6's price: $2 input, $6 output per 1M tokens

Grok 4.7 was released on September 21, 2026. The announcement calls it the company’s most powerful model for coding and knowledge work, served at the same price and speed as Grok 4.6. It is in the API as grok-4.7, and in Cursor and Grok Build. The company now signs its pages as SpaceXAI LLC.

What it costs

API prices per 1M tokens, from the pricing page on October 18, 2026:

Prompt size	Input	Cached input	Output
Under 200K tokens	$2.00	$0.50	$6.00
200K tokens and over	$4.00	$1.00	$12.00

• At 200K prompt tokens or more, the higher rate applies to all tokens in the request.
• The US regional endpoint costs 1.1x: $2.20 input and $6.60 output.
• Priority processing costs 2x. There is no batch discount for Grok 4.7.
• Grok 4.7 Fast, the same model at about twice the output speed, costs 2x the standard rates (1.5x for long prompts). It is sold only through Cursor and Grok Build, not the public API, and isn’t part of Grok Build’s free tier.

Context window

The pricing page lists a 500,000-token context window for Grok 4.7, the same as Grok 4.6 and 4.5.

Scores against Grok 4.6

Benchmark	Grok 4.7	Grok 4.6
DeepSWE v1.1	71.0%	65.2%
Terminal-Bench 4.0	37.6%	20.3%
CursorBench 4.0	46.3%	40.4%
HealthBench Professional	56.7%	48.5%

These are the company’s own numbers.

What to do

If you call grok-4.6, switching to grok-4.7 costs nothing extra per token. For long documents, keep prompts under 200K tokens where you can, since crossing that line doubles the price of the whole request. If you need more than 500K tokens of context, the older grok-4.3 takes 1M tokens at $1.25 input and $2.50 output.