Frequently asked questions about Token Plan Personal Edition quotas, purchases, subscriptions, and integration.
Quotas and limits
What do the 5-hour and 7-day limits mean?
Token Plan Personal Edition uses a dual quota system with a 5-hour rolling window and a 7-day fixed window, both measured in Credits. When cumulative consumption within a window reaches its limit, the service is paused until that window ends and the quota resets. Unused quota within a window does not carry over to the next cycle. The 5-hour limit is currently lifted for a limited time and is not enforced.
Example: The Standard plan has a 7-day limit of 10,000 Credits. You make your first call on July 20, and the system starts the window (July 20–July 27). You consume 4,000 Credits on July 20 and 6,000 Credits on July 22, reaching the 10,000 Credits limit, so the service is paused. On July 27 the window ends, the quota resets to 10,000 Credits, and the service resumes.
The limits for each tier are as follows:
Tier | 5-hour limit | 7-day limit |
|---|---|---|
Lite plan | 700 Credits | 2,500 Credits |
Standard plan | 3,000 Credits | 10,000 Credits |
Pro plan | 12,000 Credits | 40,000 Credits |
Does the 7-day limit reset on a fixed date?
No. The 7-day limit uses a fixed window mechanism. A 7-day timer starts from your first call, and the quota resets when it elapses. The reset time depends on when you first called, not a fixed calendar date (such as every Monday).
What happens when the quota is used up?
Once the limit is reached, calls are blocked and no pay-as-you-go charges apply. You can restore access by:
- Waiting for the quota to reset.
- Upgrading your plan.
- Purchasing a usage bundle for additional Credits that are not subject to window limits.
What is a reset card and how do I use it?
A reset card is a one-time quota reset entitlement for Token Plan Personal Edition subscribers. In the plan quota area of the Token Plan subscription page, click Reset limit to reset the plan quota immediately. The button shows the number of available uses (for example, 1 available), and the info icon next to it displays the tooltip "Use a reset card to reset your plan quota immediately." A reset card is granted once rather than on a recurring basis, and no separate email notification is sent when it is granted.
Why do I still see pay-as-you-go charges after activating Token Plan?
Seeing pay-as-you-go charges after activating Token Plan is usually caused by one of the following:
- Calls before activation: Calls made before you activated Token Plan are billed independently and cannot be offset by the plan.
- Configuration errors: Not using the Token Plan dedicated API Key and Base URL (for example, mistakenly using the general Model Studio endpoint dashscope.aliyuncs.com or a Coding Plan Key) routes requests through the pay-as-you-go channel.
- Unsupported models: You called a model outside the Token Plan allowlist (such as Qwen3-VL-Plus or certain sub-variants).
What is the relationship between usage bundles and plans? Do I need to subscribe to a plan first?
Usage bundles supplement a plan, providing additional Credits that are not subject to plan limits. You must subscribe to a valid plan before purchasing a usage bundle, and you can hold up to 5 bundles at the same time.
Are usage bundle Credits subject to window limits?
No. Usage bundle Credits are not subject to the plan's 5-hour or 7-day window limits and can be used immediately after purchase.
How long is a usage bundle valid?
Usage bundles are valid for 1 month. Unused Credits expire automatically when the bundle expires and are non-refundable.
Where can I view my Token Plan usage?
On the Token Plan > My Subscriptions page of the Model Studio console, view the Credits quota and consumption of your current subscription.
Why does the Credits consumption on my bill or in a plug-in differ from what I expect?
A difference between Credits consumption and what the bill or a plug-in reports is commonly caused by:
- Context cache hits: The deduction ratio is lower on a cache hit and higher on a miss.
- Model family differences: Different models (such as the Qwen series and the GLM series) have different Credits deduction ratios.
- Feature mode impact: Frequently switching tool calling or thinking mode can affect billing.
Does Token Plan offer a free trial quota or free Tokens?
The Token Plan itself does not offer a trial quota or free Tokens. Model Studio provides a separate free quota for some models; you can check the console for the list of models with available free quota.
How can I reduce Token Plan's Credits consumption?
- Compress message history or start a new conversation to avoid accumulating consumption from a long context.
- Turn off Thinking Mode to reduce reasoning overhead.
- Switch to a lighter model (such as qwen3.6-plus instead of qwen3.7-max) for less complex tasks.
- Take advantage of caching — tokens that hit the cache are billed at a lower rate than regular input tokens.
- Monitor consumption details on the Usage Analysis page and adjust your usage accordingly.
Why does the Token usage shown by a third-party tool (such as Claude Code) differ from the Model Studio console?
- Different measurement scope: Third-party tools show only model-level input and output tokens; the Model Studio console counts the full set of consumption items.
- Hidden consumption items: System prompts, tool definitions (schemas), user configuration, project conventions, accumulated multi-turn conversation history, tool call parameters and results, and internal model reasoning content all count toward Credits consumption but are not shown by third-party tools.
- Model unit price impact: Credits consumption depends on the model's deduction ratio — a higher-priced model (such as qwen3.7-max) results in higher Credits consumption for the same number of tokens.
- Optimization suggestions: View the official statistics on the Usage Analysis page of the Token Plan subscription page in the Model Studio console. Use
/compactto compress message history or/clearto start a new conversation and reduce context accumulation.
Integration errors
Common errors and solutions
Error message | Possible cause | Solution |
|---|---|---|
401 InvalidApiKey: No API-key provided. | No API Key included in the request header | Generate an API Key and configure it in your tool |
401 InvalidApiKey: Invalid API-key provided. | Using a pay-as-you-go API Key or a Coding Plan Key by mistake; subscription expired; API Key copied incompletely | Confirm you are using the Token Plan Personal Edition API Key, complete and with no extra spaces |
404 model_not_found: Model not exist. | Model name is misspelled or not in the supported list | Confirm the model name is case-sensitive and matches a model ID supported by your plan. |
403 AccessDenied.Unpurchased: Access to model denied. | The model is supported only by Token Plan Team Edition, not Personal Edition | Switch to a model supported by Personal Edition, or use Token Plan Team Edition |
401 invalid access token or token expired | Using a Coding Plan or other billing mode Base URL by mistake | Use the Token Plan Personal Edition Base URL |
401 Incorrect API key provided | Using the general Model Studio Base URL (dashscope.aliyuncs.com) by mistake | Use the Token Plan Personal Edition Base URL |
429 Requests rate limit exceeded | Too many requests in a short period | Wait one minute before retrying and reduce the request frequency |
429 Allocated quota exceeded | 5-hour or 7-day limit exhausted | Wait for the window to release quota |
400 invalid_parameter_error: Unexpected item type in content. | Non-text content such as an image was passed to a text-only model | Switch to a model that supports the modality, or confirm the request content matches the model's capabilities |
502 Bad Gateway | Intermittent server-side error | Retry later |
My Token Plan usage shows 0, or calls are still charged or generate arrears. What should I do?
- A usage of 0 usually means no consumption has occurred yet, or that an API Key other than the Token Plan dedicated Key was used.
- Calls still deducting a balance or generating arrears means a regular API Key was configured instead of the plan's dedicated Key.
Concurrency and performance
How many concurrent Agents are supported?
Concurrency depends on your plan tier:
Tier | Recommended concurrency |
|---|---|
Lite plan | Supports 1-2 concurrent Agents |
Standard plan | Supports 3-4 concurrent Agents |
Pro plan | Supports 6-8 concurrent Agents |
Will responses slow down during peak hours?
Queuing may occur during peak hours. For more stable throughput, upgrade to a higher tier or use the Team Edition.
What is Token Plan's rate limit (TPM/RPM)? Can it be increased?
The exact TPM/TPS/RPM thresholds are not publicly disclosed; rate limits are adjusted dynamically based on overall load to keep the service stable. Plan rate limits cannot be increased.
Optimization suggestions: trim your context and reduce task complexity to lower the input tokens per call; if you are rate-limited, wait about 1 minute before retrying.
Usage rules
What does "production API automation is prohibited" mean exactly?
Token Plan Personal Edition is only for individual interactive development through officially designated tools (such as Cursor, Claude Code, and Windsurf). Using the API Key for production automated services, batch scripts, scheduled background tasks, or other non-interactive scenarios is not allowed.
Can multiple people share one account?
No. Token Plan Personal Edition is limited to a single user; sharing the same account or API Key among multiple people is not allowed. For multi-user collaboration, use the Token Plan Team Edition.
Can I use the same API Key on multiple devices?
Yes. Each Token Plan Personal Edition subscription has a dedicated API Key; copy and save it securely as soon as it is generated. You can configure the same API Key on multiple devices (such as your home and office computers) without regenerating it for each device.
Purchases and subscriptions
Can RAM users use Token Plan?
Yes, but the Alibaba Cloud account must complete the following authorization:
- In the RAM console, grant the RAM user the
AliyunTokenPlanReadOnlyAccess(read-only) orAliyunTokenPlanFullAccess(management) system policy, and also grant theAliyunBSSReadOnlyAccesssystem policy. - On the Account Management page of the Model Studio console, assign the Administrator or Subscribe Plan permission to the RAM user.
Can I upgrade? How is the quota calculated after an upgrade?
Upgrading from a lower tier to a higher tier is supported. You pay the prorated price difference for the remaining period, and after the upgrade the limit is immediately raised to the quota of the new tier.
Can I downgrade?
Downgrading is not supported. To switch to a lower tier, repurchase after the current subscription expires.
How do I cancel auto-renewal?
Log on to the Model Studio console Token Plan page and disable auto-renewal in subscription management.
How do I check the effective time and remaining days of my Token Plan?
- Effective time: Log on to the Model Studio console > Token Plan page to view when the plan took effect.
- Remaining days display logic: The remaining days shown in the console are rounded down to complete 24-hour periods; the actual validity is based on the exact end timestamp (for example, 22 days and 20 hours remaining is shown as 22 days). This is expected behavior.