Pricing
Prices are in USD per million tokens. Billing is per request, metered to the token.
The more you top up, the less you pay
Your tier is applied automatically to every model below — no contract, no negotiation.
StandardYour tier
Standard pricing
Everyone starts here
VIP
10% off
Top up $100 in total
SVIP
20% off
Top up $500 in total
Tiers are permanent — we never downgrade you.
Text
| Model | Input / 1M | Output / 1M | Cached input / 1M |
|---|---|---|---|
| deepseek-ai/DeepSeek-V4-Flash977K context | $0.5280 | $1.5840 | $0.0170 |
| deepseek-ai/DeepSeek-V4-Pro977K context | $1.5840 | $4.7520 | $0.0530 |
| zai-org/GLM-5.3977K context | $1.3720 | $4.8000 | $0.3430 |
| zai-org/GLM-4.7195K context | $0.6850 | $2.7430 | $0.1370 |
| moonshotai/Kimi-K2.7-Code256K context | $1.1150 | $4.6290 | $0.2230 |
Vision
| Model | Input / 1M | Output / 1M | Cached input / 1M |
|---|---|---|---|
| deepseek-ai/DeepSeek-V4-Flash-Vision977K context | $0.5280 | $1.5840 | $0.0170 |
| zai-org/GLM-5.3-Flash977K context | $0.1370 | $0.4800 | $0.0400 |
| moonshotai/Kimi-K2.6256K context | $1.1150 | $4.6290 | $0.1880 |
Cached input applies when a request reuses a prefix the upstream has already processed. It is billed at the cached rate automatically — nothing to enable.