The PAI token service enables you to call open source and closed source large models for model evaluation, distillation, DSW, and FeatureStore, with pay-as-you-go billing based on token usage.
Billing rules
Activating the PAI Token service enables pay-as-you-go (post-paid) billing by default. You are charged based on the number of input and output tokens.
Billing formula: usage fee = number of input tokens × input unit price + number of output tokens × output unit price
Cache discount: You receive a discount on input tokens for requests that hit the context cache. For details, see Appendix: Context Cache Discount.
Text generation - Qwen
Qwen-Max
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Mode |
Input tokens per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
qwen3.7-max |
non-thinking and thinking modes |
0 < tokens ≤ 1M |
$1.98 |
$5.9412 |
International
Deployment region: Singapore.
|
Model ID |
Mode |
Input tokens per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
qwen3.7-max |
non-thinking and thinking modes |
0 < tokens ≤ 1M |
$3 |
$9 |
Qwen-Plus
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Input token range per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) |
|
|
Non-thinking mode |
Thinking mode (Chain-of-Thought + Answer) |
|||
|
qwen3.7-plus |
0 < tokens ≤ 256K |
$0.3444 |
$1.3764 |
$1.3764 |
|
256K < tokens ≤ 1M |
$1.032 |
$4.128 |
$4.128 |
|
|
qwen3.6-plus |
0 < tokens ≤ 256K |
$0.3312 |
$1.9812 |
$1.9812 |
|
256K < tokens ≤ 1M |
$1.3212 |
$7.9224 |
$7.9224 |
|
|
qwen3.5-plus |
0 < tokens ≤ 128K |
$0.138 |
$0.8256 |
$0.8256 |
|
128K < tokens ≤ 256K |
$0.3444 |
$2.064 |
$2.064 |
|
|
256K < tokens ≤ 1M |
$0.6876 |
$4.128 |
$4.128 |
|
International
Deployment region: Singapore.
|
Model ID |
Input token range per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) |
|
|
Non-thinking mode |
Thinking mode (Chain-of-Thought + Answer) |
|||
|
qwen3.7-plus |
0 < tokens ≤ 256K |
$0.48 |
$1.92 |
$1.92 |
|
256K < tokens ≤ 1M |
$1.44 |
$5.76 |
$5.76 |
|
|
qwen3.6-plus |
0 < tokens ≤ 256K |
$0.6 |
$3.6 |
$3.6 |
|
256K < tokens ≤ 1M |
$2.4 |
$7.2 |
$7.2 |
|
|
qwen3.5-plus |
0 < tokens ≤ 256K |
$0.48 |
$2.88 |
$2.88 |
|
256K < tokens ≤ 1M |
$0.6 |
$3.6 |
$3.6 |
|
Qwen-VL
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Mode |
Input tokens per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
qwen3-vl-plus |
non-thinking and thinking modes |
0 < tokens ≤ 32K |
$0.1716 |
$1.7208 |
|
32K < tokens ≤ 128K |
$0.258 |
$2.58 |
||
|
128K < tokens ≤ 256K |
$0.516 |
$5.1612 |
International
Deployment region: Singapore.
|
Model ID |
Mode |
Input tokens per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
qwen3-vl-plus |
non-thinking and thinking modes |
0 < tokens ≤ 32K |
$0.24 |
$1.92 |
|
32K < tokens ≤ 128K |
$0.36 |
$2.88 |
||
|
128K < tokens ≤ 256K |
$0.72 |
$5.76 |
Text generation - Third-party models
DeepSeek
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
deepseek-v4-pro |
$1.98 |
$3.9612 |
|
deepseek-v4-flash |
$0.1656 |
$0.33 |
International
Deployment region: Singapore.
|
Model ID |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
deepseek-v4-pro |
$2.88 |
$5.76 |
|
deepseek-v4-flash |
$0.24 |
$0.48 |
Kimi
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Input unit price (per million tokens) |
Output unit price (per million tokens) |
|
kimi-k2.7-code |
$1.0728 |
$4.4556 |
GLM
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Mode |
Input tokens per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
glm-5.2 |
non-thinking and thinking modes |
No tiers |
$1.32 |
$4.6212 |
|
glm-5.1 |
non-thinking and thinking modes |
0 < tokens ≤ 32K |
$0.99 |
$3.9612 |
|
32K < tokens ≤ 200K |
$1.32 |
$4.6212 |
||
|
glm-5 |
non-thinking and thinking modes |
0 < tokens ≤ 32K |
$0.6876 |
$3.096 |
|
32K < tokens ≤ 166K |
$1.032 |
$3.7848 |
International
Deployment region: Singapore.
|
Model ID |
Mode |
Input tokens per request |
Input unit price (per million tokens) |
Output unit price (per million tokens) Chain-of-Thought + Answer |
|
glm-5.1 |
non-thinking and thinking modes |
0 < tokens ≤ 200K |
$1.68 |
$5.28 |
Text embedding
You are charged based on input tokens only.
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Input unit price (per million tokens) |
|
text-embedding-v4 |
$0.0864 |
International
Deployment region: Singapore.
|
Model ID |
Input unit price (per million tokens) |
|
text-embedding-v4 |
$0.084 |
|
text-embedding-v3 |
$0.084 |
Multimodal embedding
You are charged based on input tokens only.
Chinese mainland
Deployment region: China (Beijing).
|
Model ID |
Input unit price (per million tokens) |
|
qwen3-vl-embedding |
Image/Video: $0.3096 Text: $0.12 |
|
multimodal-embedding-v1 |
Free Trial |
International
Deployment region: Singapore.
|
Model ID |
Input unit price (per million input tokens) |
|
tongyi-embedding-vision-plus |
$0.108 |
|
tongyi-embedding-vision-flash |
Image/Video: $0.036 Text: $0.108 |
Appendix: Context cache discounts
|
Item |
Explicit cache |
Implicit cache |
|
Token cost for cache creation |
125% of the input token price |
100% of the input token price |
|
Input token cost for cache hits |
10% of the input token price |
20% of the input token price |