| Plan |
HC / month |
| Trial | 0 |
| Starter (9€) | 1,500 |
| Plus (19€) | 6,000 |
| Pro (39€) | 20,000 |
| Enterprise (119€) | 50,000 |
Recharge: 10€ → 10,000 HC, 50€ → 50,000 HC, 100€ → 100,000 HC. Recharged HC never expire.
Economy Mode: When your HC run out, only Level 0 models (free) are available.
Bring Your Own Keys: Use your own API keys = 0 HC consumed. You pay the provider directly.
| Level |
HC/M Range |
Description |
Models |
| L0 | 0 | Free (own GPU) | 1 |
| L1 | 1–500 | Económico | 5 |
| L2 | 501–2,000 | Bajo coste | 11 |
| L3 | 2,001–8,000 | Mid-range | 20 |
| L4 | 8,001–20,000 | Premium | 10 |
| L5 | 20,001+ | Ultra | 1 |
weighted = uncached_input + (cached_input × 0.1) + (output × 5) + (reasoning × 5)
Output and reasoning tokens cost 5× more than input. Cached input costs 10× less.
| Type |
Multiplier |
| Input (uncached) | × 1 |
| Input (cached) | × 0.1 |
| Output | × 5 |
| Reasoning | × 5 |
Example: GLM-5.2 (Level 3, 5,138 HC/1M), 1,000 input + 500 output = 3,500 weighted = 6 HC