xAIReleased on Jul 9, 2025

Grok-4: API Pricing, Context Window & Benchmarks

Grok-4 is a language model from xAI, released in July 2025, with multimodal input.

Grok 4, announced by xAI in summer 2025, represents a major leap in AI capabilities, described as 'the smartest AI in the world.' Built on version 6 of xAI's foundation model, it uses 100x more training compute than Grok 2 and 10x more

Input
TextImage
Output
Text

Grok-4 benchmarks

Rankings

Quality Tracker

Grok-4 Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

LLM Stats Logollm-stats.com - Tue Jul 21 2026
Notice missing or incorrect data?

Grok-4 pricing

Providers

Grok-4 starts at $3.00 per million input tokens and $15.00 per million output tokens via xAI.

ProviderInput $/MOutput $/MContext in / outTTFT p50 / p95 sOutput avg / p5 c/sSuccess 7dModalities in / out
xAI logoxAI
$3.00$15.00256.0K/8.0K
/0.70
100/
/

Cached input is the discounted price for prompt tokens served from a provider cache. TTFT is time to first token. Output is characters per second; p5 is the sustained floor exceeded by 95% of observed requests. Success is calculated from completed versus failed requests over the trailing seven days.

Grok-4 context window

Input and output token limits for Grok-4, plus how it ranks on long-context understanding.

InputOutput
256Ktokens
8Ktokens
385 pages of text
256K
8K128K1M

Grok-4 API

Available from the model provider

Grok-4 has an official provider API. It is not currently routed through the LLM Stats gateway.

Read the official API documentation

Grok-4 latency

Grok-4 time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.

Grok-4 examples

Recent arena outputs from Grok-4, picked from the highest-ranked matchups.

Grok-4 license

Grok-4 is a proprietary model available under its provider's product and API terms, has a knowledge cutoff of December 2024.

License
Proprietary
Hosted access
Knowledge cutoff
December 2024

Proprietary license - usage restrictions apply

Grok-4 resources

Official sources for Grok-4: api documentation.

Grok-4 vs other models

The most-compared alternatives to Grok-4 are GPT-5 High, Grok-3, LongCat-Flash-Thinking. Open any pair side-by-side for benchmarks, pricing, context, and latency.

Models like Grok-4

Models ranked just above and below Grok-4 by LLM Stats score.

 

GPT-5 High

Score pending
 

Grok-3

Score pending
 

LongCat-Flash-Thinking

Score pending
 

Claude Opus 4.5

Score pending
 

ERNIE 5.0

Score pending
 

Nova 2 Pro

Score pending

FAQ

Common questions about Grok-4.

When was Grok-4 released?

Grok-4 was released on July 9, 2025 by xAI. This is the official Grok-4 release date tracked on LLM Stats.

How much does Grok-4 cost?

Grok-4 pricing starts at $3.00 per million input tokens and $15.00 per million output tokens via xAI, the lowest price among tracked providers.

Is Grok-4 available via API?

Yes, Grok-4 is available via API. See the official documentation for authentication and endpoint details. It is served by 1 provider tracked on LLM Stats.

Who created Grok-4?

Grok-4 was created by xAI.

What is the license for Grok-4?

Grok-4 is released under the Proprietary license.

What is the knowledge cutoff date for Grok-4?

Grok-4 has a knowledge cutoff of December 2024, meaning it was trained on data up to that point and may not know about events after it.

Is Grok-4 multimodal?

Yes, Grok-4 is multimodal and can accept both text and images as input.

What is Grok-4 latency?

Grok-4 p95 time to first token is 0.70 seconds via xAI over the trailing 7 days. Lower time to first token means the model begins responding sooner for chat, agents and API workloads.

Where can I use Grok-4?

Grok-4 is available through 1 provider including xAI.

What models should I compare Grok-4 against?

Common Grok-4 comparisons include Grok-4 vs GPT-5 High, Grok-4 vs Grok-3, Grok-4 vs LongCat-Flash-Thinking. Compare them side by side for benchmark scores, pricing, context window, latency and API availability.