About Multipliers#
Definition: Multipliers (group multipliers, model multipliers, and completion multipliers) are a system calculation method. Each model has its own unique multiplier and does not directly reflect the official price multiplier.
Principle: In theory, our multipliers are designed to be completely consistent with official model quotes.
About Price Adjustments#
GPT Models: Generally synchronized with official multipliers unless otherwise announced.
Other Manufacturers' Models: Multiplier prices are subject to market fluctuations.
You can subscribe to email notifications about rate changes through the Subscription Page. About Concurrency#
This site uses large-scale cluster load balancing, supporting ultra-high and ultra-large concurrency by default.
For ultra-high concurrency requirements or specific model training needs, we recommend using training and contact customer service to activate the relevant model permissions.
Enterprise Dedicated Line: Supports ultra-high concurrency of 3000W TPM by default, and is supported by all-inclusive lines.
Request Volume Monitoring: Visit the Request Log Page to observe changes in RPM request volume. About Maintenance#
Multi-Channel Distribution:
When a channel fails, the system automatically switches to another channel and retries multiple times. - If the channel crashes, the system will automatically restart within 5 minutes.
If the model returns an error message in a high-concurrency scenario, please contact customer service for assistance.
403 Error: This may have triggered the WAF defense mechanism. Please contact customer service to check if the URL is blocked.
Scheduled Maintenance Time:
1-5 AM: Executing scheduled tasks, database maintenance, freeing up memory, etc., may cause brief instability or interruptions.
Recommendation: Users making batch concurrent requests should implement a reconnection mechanism and save request data.
Privacy Statement#
Logged Information: To ensure service quality and network security, we will record users' request IP addresses, request frequency, and service credits used.
Aside from necessary service log information, we do not collect any other personal request data.
User request content is not stored to ensure the utmost respect and protection of user privacy.
Top-up Discounts#
Rules: The more you top up online, the greater the discount. Transparent and no hidden fees.
We will never use "super low prices" or "multipliers x times higher than the official website" scams.
Cumulative top-ups of ¥3,000 or above or daily average spending of over ¥100. For corporate customers, please contact customer service to enjoy a permanent top-up or spending discount.
Regarding Abnormal Charges#
1.
Check the Requesting IP in the Request Log.
If it's an unfamiliar IP, your key may have been leaked.
Disable or delete the leaked token and create a new one.
Token Protection: Do not share the token casually or enter it on unauthorized websites.
Notes on Open Source Programs:
If using a personally built open source AI program, please ensure:
Set a password for the site.
Do not save tokens via environment variables; instead, use form submission to save.
Modify the header and footer signature information in open-source programs to prevent malicious scanning and theft.
About Deposit Arrivals#
Delay Note: Deposit processing may be delayed by 1-3 minutes on some networks; please be patient.
If your funds have not been received for more than 10 minutes:
1.
Contact customer service and provide a screenshot of your payment and your order number.
2.
We will manually process your order.
Modified at 2025-10-15 14:07:20
