OpenAI · live
The cheapest way to use the GPT API
A GPT token costs the same list price wherever it's billed. You save in two places: the host you route to, and what you pay for the credits.
Latest GPT models, list price
| Model | Input / 1M | Output / 1M | Cached input / 1M | Context |
|---|---|---|---|---|
GPT-6.2 Astra Pro openai/gpt-6.2-astra-pro | $8.00 | $40.00 | $0.4 | 1,050K |
GPT-6.2 Astra openai/gpt-6.2-astra | $8.00 | $40.00 | $0.4 | 1,050K |
GPT-6.2 Sol Pro openai/gpt-6.2-sol-pro | $2.00 | $10.00 | $0.1 | 1,050K |
GPT-6.2 Sol openai/gpt-6.2-sol | $2.00 | $10.00 | $0.1 | 1,050K |
GPT-6.1 Terra Pro openai/gpt-6.1-terra-pro | $1.00 | $5.00 | $0.05 | 1,050K |
GPT-6.1 Terra openai/gpt-6.1-terra | $1.00 | $5.00 | $0.05 | 1,050K |
GPT-6.1 Sol Pro openai/gpt-6.1-sol-pro | $2.00 | $10.00 | $0.1 | 1,050K |
GPT-6.1 Sol openai/gpt-6.1-sol | $2.00 | $10.00 | $0.1 | 1,050K |
GPT-6 Luna Pro openai/gpt-6-luna-pro | $0.1 | $0.5 | $0.01 | 1,050K |
GPT-6 Luna openai/gpt-6-luna | $0.1 | $0.5 | $0.01 | 1,050K |
From OpenRouter's public model list, which passes OpenAI's prices through without markup. Refreshed hourly.
Same model, different hosts
Some models are served by several clouds. Prices usually match the list price; when one host is cheaper, routing to it saves money on every token.
GPT-6.2 Astra Pro
- OpenAI$8.00 / $40.00
- Azure$8.00 / $40.00
GPT-6.2 Astra
- OpenAI$8.00 / $40.00
- Azure$8.00 / $40.00
- Amazon Bedrock$8.00 / $40.00
GPT-6.2 Sol Pro
- OpenAI$2.00 / $10.00
- Azure$2.00 / $10.00
GPT-6.2 Sol
- OpenAI$2.00 / $10.00
- Azure$2.00 / $10.00
- Amazon Bedrock$2.00 / $10.00
Where GPT usage costs least to buy
| Usage bought → | $20 | $100 | $500 | $1,000 | $2,000 |
|---|---|---|---|---|---|
| Orbio · Card | $11.50 | $53.50 | $263.50 | $526.00 | $1,072.56 |
| Direct from the lab · Card | $20.00 | $100.00 | $500.00 | $1,000.00 | $2,000.00 |
| OpenRouter · Crypto | $21.00 | $105.00 | $525.00 | $1,050.00 | $2,100.00 |
| OpenRouter · Card | $21.10 | $105.50 | $527.50 | $1,055.00 | $2,110.00 |
Cheapest per column highlighted. * card processing and tax added at checkout, not in the number. Live quotes, refreshed every 5 minutes.
"Direct" is prepaid credit from OpenAI at face value (platform.openai.com). The others reach GPT through OpenRouter. Cut the token bill itself with batch and caching.