Memory & Model rules
Unlike a purely pay-as-you-go system, we use a hybrid pricing model designed to keep short chats simple and predictable while ensuring fair pricing for longer conversations.
Your Coin cost depends on your selected tier and Current Memory Usage:
Up to 8,000 tokens
- A fixed Coin price applies.
Above 8,000 tokens
- Coins are charged based on your exact Current Memory Usage.
What Current Memory Usage Includes
- Current Memory Usage includes your active conversation, character background, pinned memories, and the bot’s replies.
Pricing Reference Table
| Tier | Up to 8 000 tokens | Above 8 000 tokens |
|---|---|---|
| Top | 96 Coins fixed | 120 Coins per 10 000 tokens |
| High | 64 Coins fixed | 80 Coins per 10 000 tokens |
| Medium | 32 Coins fixed | 40 Coins per 10 000 tokens |
| Low | 16 Coins fixed | 20 Coins per 10 000 tokens |
Formula
When Current Memory Usage is 8 000 tokens or less:
Coins charged = Fixed price for your selected tier
When Current Memory Usage is above 8 000 tokens:
Coins charged = Current Memory Usage ÷ 10 000 × Tier Rate
Examples
Scenario A: Short Chat
Your Current Memory Usage is 5 000 tokens.
| Tier | Coins charged |
|---|---|
| High | 64 Coins |
| Medium | 32 Coins |
| Low | 16 Coins |
Scenario B: Long Chat
Your Current Memory Usage is 12 000 tokens.
| Tier | Calculation | Coins charged |
|---|---|---|
| High | 12 000 ÷ 10 000 × 80 | 96 Coins |
| Medium | 12 000 ÷ 10 000 × 40 | 48 Coins |
| Low | 12 000 ÷ 10 000 × 20 | 24 Coins |