Billing & pricing

​

Billing options

​

What's the difference between pay-as-you-go and Coding Plan?

Pay-as-you-go (API usage)

• Billed per million tokens consumed
• Input and output priced separately
• Best for: API integrations, applications, batch processing
• Flexible: Pay only for what you use
• Supports all optimization features (batch API, context cache)

Coding Plan (subscription)

• Fixed monthly fee ($50/month)
• 90,000 requests per month
• Best for: AI coding tools (Claude Code, Cursor, Cline)
• Predictable cost for daily coding workflows
• Uses separate API key (sk-sp-xxxxx)

​

Which billing option should I choose?

• Building an app or API integration? → Use pay-as-you-go
• Using AI coding tools daily? → Use Coding Plan
• Just getting started? → Start with pay-as-you-go and free tokens

​

Bill analysis

​

How do I analyze bill details?

Qwen Cloud bills support granular reconciliation by API Key ID, workspace ID, model name, input/output type, and invocation channel.

1. On the Billing Overview page, select a billing month.
2. Set Product Name to Qwen Cloud and click Search.
3. Click Export (upper-right) to download the bill.
4. Open the file and locate the Instance ID column.

Instance ID contains multiple values separated by semicolons (;):

• Example: text_token;llm-xxx;qwen-max;output_token;app
• Format: Billing Type;Workspace ID;Model Name;Input/Output Type;Call Channel

Look up each field:

• API key: Copy the API Key ID from the bill and find it on the API Key Management page.
• Workspace: Copy the workspace ID from the bill and find it on the Workspaces page.
• Call channel:
  • app: Model called via application code
  • bmp: Model called from Try AI
  • assistant-api: Model called via Assistant API

​

Why can't I find my bill immediately after calling a model?

Possible reasons:

1. Bill delay: Bills update hourly but may lag during peak hours (costs from 16:00-17:00 might appear at 19:30).
2. Free tier: No charges or bills are generated for usage within free quota or when calling free models.

​

Where can I view model call statistics?

Check the Model Observation page.

​

Payment and charges

​

Is pay-as-you-go billed in real time?

No. The system freezes credit upfront and settles at month end (early the following month).

​

How do I stop billing?

• Stop using the Playground or making API calls.
• Delete API keys on the API Key page to prevent accidental calls.

​

Do built-in tools like web search cost extra?

Yes. Some built-in tools charge a per-call fee on top of model token costs. Web search costs $10 per 1,000 calls; image search costs$8 per 1,000 calls. Web extractor and code interpreter are temporarily free. See Pricing for the full table. Tool call fees are not covered by free token quota.