GPT vs Claude API Cost Comparison
Compare selected OpenAI and Anthropic models against the same standard text workload. The estimate shows token charges, not quality or full project costs.
Prices checked October 17, 2026. The first selector stays within OpenAI, and the second stays within Anthropic. Both use the same request and token assumptions.
Workload assumptions
One request means one call to the selected model, across all users.
Include reasoning or thinking tokens if the provider bills them as output. This estimate does not infer them.
Above 272,000 input tokens, the higher rate applies to the full request.
Official model pricingCost as request volume grows
The same input and output tokens per request apply to both models.
| Total requests / day | Requests / month | GPT-6.1 Sol / mo | GPT-6.1 Sol / year | Claude Sonnet 5.5 / mo | Claude Sonnet 5.5 / year |
|---|---|---|---|---|---|
| 100 | 3.0K | $21.00 | $255.50 | $21.00 | $255.50 |
| 1.0K | 30.0K | $210.00 | $2,555.00 | $210.00 | $2,555.00 |
| 10.0K | 300.0K | $2,100.00 | $25,550.00 | $2,100.00 | $25,550.00 |
| 100.0K | 3.0M | $21,000.00 | $255.5K | $21,000.00 | $255.5K |
Monthly estimates use 30 days. Annual estimates use 365 days. The estimate covers standard text input and output only. It excludes cached-token pricing, batch discounts, external tools and search, embeddings, vector storage, hosting and infrastructure, negotiated discounts, free-tier credits, and image, audio, video, or voice usage.
