Model rates. Clearly stated.
Published customer prices in USD per million tokens. Compare input, output and eligible cached input before making a request.
Compare published rates.
USD per million tokens. The effective request limits and supported fields are documented on each model page.
Scroll sideways to see all rates and limits.
| Model | Input | Output | Cached input | Context |
|---|---|---|---|---|
| NotHoly SpecterGLM 5.3 Abliterated | $3.00 | $6.00 | $0.45 | 1,048,576 |
| NotHoly Pulse FlashGLM 5.3 Flash Abliterated | $0.36 | $1.50 | $0.09 | 1,048,576 |
| NotHoly Vector FlashDeepSeek V4.1 Flash 1017 Abliterated | $0.42 | $1.50 | $0.105 | 1,048,576 |
Estimate a request
Use the same token counts to compare published customer rates.
Cached tokens are part of the total input, charged at the cached rate instead of the ordinary input rate.
Estimates use the published rates and request limits. Actual billable context, output and eligible cache hits determine final usage.
From tokens to credits.
100 credits represent $1 of usage at the published retail rates. Fractional credits are preserved. A credit is a usage unit, not withdrawable money or a fixed number of messages.
Cached input replaces ordinary input for the eligible subset; it is not billed twice. Each model page shows the rate schedule and current effective limits.
Control what you spend.
Set an output limit and a customer-key spending limit. Review the current credit and purchase terms in billing before checkout. Automatic top-ups are disabled by default.
Tool operations have separate explicit budgets. Model usage does not authorize unlimited scanners, compute or storage.