Understanding Model Pricing
Every request is billed per token. Input tokens are what you send, output tokens are what the model replies with.
Why output costs more
Generating text is more expensive than reading it. Most providers price output tokens higher than input tokens.
A practical example
| Request | Input tokens | Output tokens | Relative cost |
|---|---|---|---|
| Short chat | 100 | 50 | Low |
| Summarize a page | 2,000 | 300 | Medium |
| Write a code module | 1,000 | 1,500 | High |
Credits vs dollars 💱
Your balance is measured in credits, which map 1:1 to the platform currency. The panel shows the exact credit cost of each model before you send a request.
Reduce the bill
- Keep system prompts short
- Use output limits
- Pick mini models for easy tasks
Frequently asked questions ❓
What exactly is a token?
A token is a chunk of text the model reads or writes — roughly four characters for English.
Why does output cost more than input?
Generating new text is more compute-intensive than reading your prompt, so output tokens are priced higher.
Where can I see per-model prices?
The panel shows the exact credit cost of each model before you send a request.
How can I lower my bill?
Shorten prompts, cap output tokens and pick mini models for simple tasks.
Comments (0)