Models with a published cache-read price
Verified cache-read tariffs. Eligibility, cache writes and hit rates differ; the read tariff alone is not a total request cost.
54 models · Page 2 of 2 · Published snapshotAnthropic API model pricesGoogle API model pricesOpenAI API model pricesxAI API model pricesModels with at least 200k contextModels with published batch prices
| Model | Cache read / 1M | Standard input / output | Context | Verified |
|---|---|---|---|---|
| $0.50 | $2.00 / $6.00 | 500,000 | ||
| $0.50 | $2.00 / $8.00 | 1,047,576 | ||
| $0.50 | $2.00 / $8.00 | 200,000 | ||
| $0.50 | $5.00 / $25.00 | 200,000 | ||
| $0.50 | $5.00 / $25.00 | 1,000,000 | ||
| $0.50 | $5.00 / $25.00 | 1,000,000 | ||
| $0.50 | $5.00 / $25.00 | 1,000,000 | ||
| $0.50 | $5.00 / $25.00 | 1,000,000 | ||
| $0.50 | $5.00 / $30.00 | 1,050,000 | ||
| $0.55 | $1.10 / $4.40 | 200,000 | ||
| $1.00 | $10.00 / $50.00 | 1,000,000 | ||
| $1.00 | $10.00 / $50.00 | 1,000,000 | ||
| $1.00 | $10.00 / $50.00 | 922,000 | ||
| $1.25 | $2.50 / $10.00 | 128,000 |
Prices are direct-provider USD tariffs per 1M tokens, observed by this registry. The standard and batch blends use 75% input and 25% output tokens. Cache-read prices exclude writes and require eligible reuse. Each model record links its official source and dated snapshot. Read the method.