EN ▾
Get your key

Transparent Pricing: Pay Only for Tokens Used

You pay strictly for tokens processed in your requests. Errors cost nothing, and your prepaid balance never expires.

$0.15
per 1M input tokens
$0.50
Output tokens / 1M
100,000
token context
$0.50
trial credit
300
requests per minute

Usage examples

  1. Long context summarization

    Input: 80,000 tokens. Output: 2,000 tokens. Cost: (80,000 / 1,000,000 * $0.15) + (2,000 / 1,000,000 * $0.50) = $0.012 + $0.001 = $0.013. You only pay for the actual tokens sent and received.

  2. Standard roleplay turn

    Input: 5,000 tokens. Output: 1,000 tokens. Cost: (5,000 / 1,000,000 * $0.15) + (1,000 / 1,000,000 * $0.50) = $0.00075 + $0.00050 = $0.00125. Short exchanges are extremely cheap.

  3. Batch processing errors

    Input: 10,000 tokens. Output: 0 tokens (API error). Cost: (10,000 / 1,000,000 * $0.15) + 0 = $0.0015. If the model refuses or errors, you still pay for the input tokens but never for the output.

Transparent Token Pricing Breakdown

Our pricing model charges only for the tokens your application processes. We do not hide fees behind monthly subscriptions or tiered limits. The base rate is $0.15 per 1 million input tokens and $0.50 per 1 million output tokens. This structure ensures that your costs scale exactly with your usage volume. Short conversations cost fractions of a cent, while heavy batch processing remains predictable. We calculate costs based on real token counts, so you never overpay for unused capacity.

Prepaid Credit & Expiration Policy

You fund your account with prepaid credit that never expires. This eliminates the risk of losing money to monthly billing cycles or auto-renewal traps. Your balance remains available indefinitely until you decide to stop using the service. Errors and content refusals are free, meaning you only pay for successful token generation. This approach keeps your operational spend predictable and aligned with actual API activity. You can top up whenever your balance runs low without any administrative delay.

Why Prepaid Credit Works Best for Developers

A prepaid balance prevents unexpected overages since you can only spend what you have loaded. This is ideal for development environments where usage might spike or drop unexpectedly. Your credit sits in your account until you need it, providing flexibility for both small experiments and large-scale deployments. The model is simple: load funds, send requests, and watch your balance decrease by the exact token cost.

Frequently asked questions

Do I lose my prepaid credit if I stop using the API?

No, your prepaid credit never expires. You can top up once and use the funds over months or years without any time limit. This ensures you only pay for the capacity you actually need.

Do I pay for tokens if the model refuses my prompt?

You pay for the input tokens sent to the API, but you do not pay for the output tokens if the response is empty or refused. Errors in processing also do not incur output charges. You only pay for the actual tokens processed.

What happens if I overpay or double charge?

Your credit remains in your account indefinitely until you request a refund through the Support page. Mistakes are resolved by adjusting your balance directly. There are no complex refund cycles or waiting periods for standard errors.

Is there a monthly fee for keeping my account active?

No, there are no monthly fees or subscription costs. You only pay for the tokens you use from your prepaid balance. Your account stays open as long as you have credit or choose to keep it active.

Generate your key now

Sign up, copy the key, update your base URL. That is the entire configuration.

Get your key