Transparent prices
Standard and cached-token prices are explicit and versioned. A price change never rewrites past usage.
Multi-model prepaid API
One API. Many models. Prepaid, never invoiced.
Use GPT, Claude, Grok, DeepSeek and GLM through one account. Point your existing [OI] client at one base URL, swap the key and the model name, and see exactly what every request costs.
Production Botsk-japi-live-a9…
Account balance
$15.00
$5.42 used$9.58 available
Cached ratefrom $0.009 / 1M
SettlementPer request
Built for control
Simple outside, strict accounting inside.
Prepaid by design
No surprise invoice: a request can only spend credit that is already on the account.
Concurrent-safe
Spending capacity is reserved before the upstream call, so parallel requests cannot overspend.
One integration
Keep one base URL while switching between permitted model names.