Skip to main content
Billing is per-use: each request is charged at the model’s rate, separately for input and output tokens. Your balance never expires, and there are no subscriptions or auto-charges.

Model rates

A single price for all models — per 1M tokens (₽):
ModelIDInput (1M)Output (1M)
Opus 4.8claude-opus-4-837 ₽185 ₽
Sonnet 4.6claude-sonnet-4-637 ₽185 ₽
Haiku 4.5claude-haiku-4-537 ₽185 ₽
Request cost:
price = (input_tokens × input_rate + output_tokens × output_rate) / 1 000 000
Opus — for the most complex tasks, Sonnet — a balance of price and quality (the default for code), Haiku — the cheapest and fastest. You can mix them per task.

Token packs

You can top up any amount, or pick a ready-made pack:
VolumePrice₽ per 1M
5M199 ₽39.8
15M559 ₽37.3
35M1 300 ₽37.1
100M3 000 ₽30.0
200M5 500 ₽27.5
300M8 000 ₽26.7
1B13 000 ₽13.0
2B22 000 ₽11.0
10B80 000 ₽8.0
The larger the volume, the cheaper 1M becomes. Prices are approximate; the current ones are in your dashboard.

Payment

  • SBP (QR) and cryptocurrency — no cards and no VPN.
  • Funds are credited automatically once the payment is confirmed.
  • Transaction history and current balance are in your dashboard.
See Balance and top-up.