GPT-5 vs Grok 4

API pricing, context window and capabilities for GPT-5 and Grok 4, side by side. Prices are the cheapest serving route per 1M tokens, verified against LiteLLM and OpenRouter.

Verdict

computed from the published numbers

GPT-5 wins on price, on intelligence and on context window. Benchmark data is unavailable for one or both models.

Price
GPT-5
Coding
Not called
Intelligence
GPT-5
Context window
GPT-5

Key takeaways

who wins on what, from the published numbers

GPT-5 wins

  • Cheaper input ($1.25 vs $3.00 /1M)
  • Cheaper output ($10.00 vs $15.00 /1M)
  • Larger context window (410K vs 256K)
  • More serving providers (6 vs 5)
  • Supports vision

Grok 4 wins

  • Higher max output (128K vs 256K)
Cheaper overall
GPT-5
$1.25 + $10.00 /1M
Larger context
GPT-5
410K
More providers
GPT-5
6 providers

Pricing & specs

cheapest serving route · per 1M tokens
GPT-5Grok 4
Input $/1M$1.25 best$3.00
Output $/1M$10.00 best$15.00
Context window410K best256K
Max output128K 256K best
Serving providers6 best5
Released2025-08-07 2024-07-09

Benchmarks

Artificial Analysis indices, side by side
GPT-5Grok 4
Intelligence34.7 best33.3
Coding37.8 Not available
GPQA Diamond85% 88% best
Output speed69 t/s Not available
Latency (TTFT)33.91s Not available

Cost calculator

choose a serving route for each side
Input tokens / month
Output tokens / month
GPT-5
via Databricks
$16
Grok 4
via Azure
$30

Capabilities

feature support, side by side
FeatureGPT-5Grok 4
Vision
Tool use
Reasoning
Audio
Prompt caching
Structured output
Web search

Related comparisons

Frequently asked questions

Which is cheaper, GPT-5 or Grok 4?
GPT-5 is cheaper for input ($1.25 vs $3.00 per 1M tokens). GPT-5 is cheaper for output ($10.00 vs $15.00 per 1M tokens).
What is the context window difference between GPT-5 and Grok 4?
GPT-5 has a 410K-token context window and Grok 4 has 256K tokens, so GPT-5 can take a larger prompt.
How do GPT-5 and Grok 4 compare on benchmarks?
On the tracked Artificial Analysis indices, GPT-5 leads on intelligence.
Pricing verified against LiteLLM + OpenRouter + provider pages · refreshed daily

Every weekday

AI moves fast. Here's your debrief.

News, analysis, tools, and more.

For people who build with AI