GPT API cost calculator
Enter your monthly usage and cache hit rate to see what the same GPT workload costs on the OpenAI API and on Wokey. Wokey prices come from the live catalog, the same data billing uses.
Monthly cost
| Model | Coding agent, daily use · OpenAI API | Coding agent, daily use · Wokey | Coding agent, light use · OpenAI API | Coding agent, light use · Wokey | Coding agent, heavy use · OpenAI API | Coding agent, heavy use · Wokey | API app, no caching · OpenAI API | API app, no caching · Wokey |
|---|---|---|---|---|---|---|---|---|
| GPT-6.1 Sol | $51.50 | $4.64 | $12.88 | $1.16 | $206.00 | $18.54 | $90.00 | $8.10 |
| GPT-5.6 Sol | $121.00 | $13.31 | $30.25 | $3.33 | $484.00 | $53.24 | $180.00 | $19.80 |
| GPT-5.6 Terra | $64.50 | $6.45 | $16.13 | $1.61 | $258.00 | $25.80 | $100.00 | $10.00 |
| GPT-5.6 Luna | $6.45 | $3.87 | $1.61 | $0.968 | $25.80 | $15.48 | $10.00 | $6.00 |
| GPT-6 Astra | $302.50 | $16.64 | $75.63 | $4.16 | $1,210.00 | $66.55 | $450.00 | $24.75 |
| GPT-6 Sol | $60.50 | $5.44 | $15.13 | $1.36 | $242.00 | $21.78 | $90.00 | $8.10 |
| GPT-6 Luna | $3.03 | $2.72 | $0.756 | $0.681 | $12.10 | $10.89 | $4.50 | $4.05 |
| GPT-6.1 Terra | $25.75 | $2.58 | $6.44 | $0.644 | $103.00 | $10.30 | $45.00 | $4.50 |
- Coding agent, daily use: A few hours of coding each workday
- Coding agent, light use: A few sessions a week, about a quarter of daily use
- Coding agent, heavy use: Parallel sessions or subagents, about four times daily use
- API app, no caching: Every request carries a different context
GPT model prices
| Model | Uncached input | Cache reads | Output |
|---|---|---|---|
| GPT-6.1 Sol | $0.18 / $2 | $0.009 / $0.1 | $0.9 / $10 |
| GPT-5.6 Sol | $0.44 / $4 | $0.044 / $0.4 | $2.2 / $20 |
| GPT-5.6 Terra | $0.2 / $2 | $0.02 / $0.2 | $1.2 / $12 |
| GPT-5.6 Luna | $0.12 / $0.2 | $0.012 / $0.02 | $0.72 / $1.2 |
| GPT-6 Astra | $0.55 / $10 | $0.055 / $1 | $2.75 / $50 |
| GPT-6 Sol | $0.18 / $2 | $0.018 / $0.2 | $0.9 / $10 |
| GPT-6 Luna | $0.09 / $0.1 | $0.009 / $0.01 | $0.45 / $0.5 |
| GPT-6.1 Terra | $0.1 / $1 | $0.005 / $0.05 | $0.5 / $5 |
How it is calculated
- Cache reads bill at the cache-read price, the rest of the input at the plain input price, and output at the output price. Where a model lists a cache-write price, written tokens bill at it; otherwise they bill as plain input.
- A coding agent resends a long context every turn, so most input is cache reads. The profiles are assumptions for estimating, not measured bills.
- The OpenAI API is priced at its published standard rate, without batch discounts, priority tiers or long-context surcharges where they apply.
- The OpenAI pricing page lists 30m and 12h cache-write prices; this page uses the 30m price, and the 12h price is twice that.
- Wokey settles each request at the price in effect when it is processed, and every call lists its metered usage and cost in the console request log.
Price sources
FAQ
How much does the GPT-6.1 Sol API cost per month?
For the "Coding agent, daily use" profile (100M input tokens (90% cache reads) and 2M output tokens): about $51.50 on the OpenAI API and about $4.64 on Wokey. The profile is an assumed month, not a measured bill; replace the numbers above with your own usage to recalculate.
How are GPT cache hits billed?
Input served from cache bills at the cache-read price, the rest at the plain input price. For GPT-6.1 Sol on Wokey that is $0.18 for input and $0.009 for cache reads per 1M tokens, so a higher cache hit rate makes the input side cheaper.
Are the Wokey prices here what I am actually charged?
Yes. Prices come from the public pricing endpoint, the same catalog billing uses, and every call lists its metered usage and cost in the console request log.