This topic describes how to query bill details, analyze bills, and stop billing.
Query bills
1. Query inference costs for a specific model
To query the inference costs for a specific model, such as qwen-plus:
On the Bill Details page, select a Billing Month.
Set Commodity Name to Model Studio Foundation Model Inference, and then click Search.
In the Instance ID column, find all instances related to qwen-plus.
Add the Amount After Price Reduction for these instances. The sum is the total inference cost for the model in the selected month.
2. Query total costs for Model Studio services
To view the overall spending trend for Model Studio or a specific service:
On the Cost Analysis page, set Cost Type to Pretax Amount.
Set Time Unit to Month and select a time range, such as May to October 2025.
Set Product Detail to Model Studio Foundation Model Inference to view the total cost of model inference for the selected time range.
On the Cost Analysis page, set Cost Type to Pretax Amount.
Set Time Unit to Month and select a time range, such as October 2025.
Set Product to Alibaba Cloud Model Studio to view the costs for Model Studio in the selected time range.
3. Query Token consumption from detailed bills
On the Bill Details page, set Statistic Item to Billable Item and export the bill. You can view the Token usage in the exported bill.
Analyze bill details
Starting September 7, 2024, bills in Model Studio support more granular reconciliation. You can view your fees by ApiKeyID, workspace ID, model name, input/output type, and invocation channel.
1. Download the bill
On the Detailed Bill page, select a Billing Month.
Set Product Name to Alibaba Cloud Model Studio, and then click Search.
In the upper-right corner of the bill list, click the Export icon to download the bill.
Open the file, find the Instance ID column, and check it against the rules described below.
2. Interpret key fields
Instance ID field: This field contains multiple pieces of information separated by semicolons (;).
Example:
text_token;llm-xxx;qwen-max;output_token;appThis represents
Billing Type;Workspace ID;Model Name;Input/Output Type;Call Channel.
3. Data tracing and term descriptions
Query an API key: Copy the
ApiKeyIDfrom the bill and go to the Model Studio Key Management page to find the corresponding key name.Query the workspace: Copy the
Workspace IDfrom the bill and go to the Workspaces page to identify the specific workspace.Call channel descriptions:
app: The model was called through an application (code).bmp: The model was called from the Playground in the console.assistant-api: The model was called through the Assistant API.
Stop billing (shut down service)
If you no longer need to use Model Studio services, stop the billing to avoid additional charges.
1. Stop model inference
Action: Charges are no longer generated after you stop using related features, such as the Playground in the Model Studio console or API calls in your code.
Recommendation: To prevent accidental calls, delete the API key on the API-Key (Singapore), API-Key (Virginia), or API-Key (Beijing) page.
2. Stop model training
Action: No charges are generated if no training tasks are in progress.
FAQ
Q: Why can't I find the bill right after calling a model?
A: The possible reasons are as follows:
Bill generation delay: Bills are updated hourly. During peak hours, the update may be delayed by hours. For example, costs incurred between 16:00 and 17:00 might not appear on the bill until 19:30.
Your free quota has not been used up: No charges are incurred if your usage is within the free quota (Singapore region only). You are charged only for usage that exceeds the free quota.
Used a free model: If you call a free model, no bill is generated.
Q: Is pay-as-you-go service billed in real time?
A: No, it is not. Alibaba Cloud's pay-as-you-go services typically use a reservation and monthly settlement model. The system first freezes a portion of your credit. At the end of each monthly billing cycle, which is early the next month, the final bill is generated and the actual deduction is made.
Q: Where can I see model call counts and statistics?
A: You can query model call statistics on the Model Observation (Singapore), Model Observation (Virginia), or Model Observation (Beijing) page.