Side by side pricing and context for Grok 4 and GPT-5. Rates are curated catalog values with last verified dates on each model page.
Grok 4
xAI frontier Grok for general chat and reasoning-style apps. Use when you want the Grok stack; confirm rates on xAI docs as the lineup moves quickly.
Input $3.00 / Output $15.00 per 1M
Context 256K · Approx
GPT-5
OpenAI frontier model for hard coding, agents, and long-context production work. Use when quality matters more than the cheapest token rate, then check cache and batch in TokenCALC before you scale.
Input $1.25 / Output $10.00 per 1M
Context 400K · Exact (o200k_base)
Quick take
- GPT-5 is cheaper on input in this catalog snapshot.
- GPT-5 is cheaper on output in this catalog snapshot.
- Both last verified on their model pages (2026-08-02 / 2026-08-02).
Rate table
| Metric | Grok 4 | GPT-5 |
|---|---|---|
| Input / 1M | $3.00 | $1.25 |
| Output / 1M | $15.00 | $10.00 |
| Cached input / 1M | n/a | $0.125 |
| Context | 256K | 400K |
| Tokenizer | Approx | Exact (o200k_base) |