Pay Only for What You Use
At Pistachio, we believe in transparent, fair pricing that scales with your needs. No hidden fees, no monthly subscriptions, no surprises—just straightforward pay-as-you-go pricing that puts you in control.
The cost of the AI model is based on the number of input tokens and output tokens. Input tokens include your prompt, images, and the encoded design sent to the model. Output tokens are used to reply to your message and modify the design. You can find these numbers for each chat at the top of the chat window. In some cases, you may find that the total number of tokens billed is higher than the tokens shown in the chat window. This is because:
- Some tokens are used for setting up the design rules you choose in the design presets window.
- Some inputs to the model are cached and charged separately.
- When you stop the model response, the model will be interrupted and you are still billed for the tokens generated that have not yet been shown in the chat.
In a long conversation, each new message is sent to the model along with the entire chat history, including token-heavy items like images. This causes token usage to grow exponentially. We use aggressive caching technology to reduce the cost of feeding the chat history to the model many times. This can reduce the cost of input tokens by 90%, and these savings are passed on to you. You can choose from two models: a fast and cost-effective model for general use cases and a more powerful thinking model for complex tasks. The following table shows the pricing for each model:
| Model | $ per million input tokens | $ per million output tokens | $ per million cached tokens | $ per million cache writes |
|---|---|---|---|---|
| Fast | $0.5 | $3 | $0.01 | $0.05 |
| Thinking | $3 | $15 | $0.3 | $3.75 |
You can view your daily token usage, costs, and credit balance in your dashboard.
The code generation feature for mobile apps is currently being developed. Exporting app bundles is free and won't cost you any credits.