Claude Opus 4.6 vs GPT-5
Sample workload: short support reply (~512 output tokens). Open the calculator for your own prompt.
| Field | Claude Opus 4.6 | GPT-5 |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Accuracy | Approx | Exact |
| Input / 1M | $5.00/1M | $1.25/1M |
| Output / 1M | $25.00/1M | $10.00/1M |
| Sample input tokens | 25 | 16 |
| Sample request cost | $0.012925 | $0.00514 |
| Pages | Model page | Model page |
Try your own prompt
Sample costs above use a fixed support-chat prompt. For your text, open the main calculator with either model preselected.
Claude Opus 4.6 in calculator · GPT-5 in calculator · Count tokens (Claude Opus 4.6) · Count tokens (GPT-5) · All comparisons
How LLM pricing works · Exact vs Approx · Claude Opus 4.6 page · GPT-5 page