Event Update Notes

Due to a business restructuring, effective immediately, the "Flagship Model Value Subscription" and the "Lightweight All-in-One Subscription" have been officially merged into the "Value Resource Package," and their pricing has been standardized. If you have previously purchased either of these subscription products, your existing benefits will remain fully intact and will not be affected in any way; please continue to use them with confidence. If you have any questions regarding this change, please feel free to contact customer service at any time.

Fusion Token Documentation

The tokens included in the subscription package are "fusion tokens," which can be used to call any model included in the package. The specific consumption rate is determined by real-time calculations at the time of the API call.

• Benchmark Definition: The benchmark price is based on the input token price of the claude-haiku-4-5-20251001 model, which is $1 per Mtoken (i.e., $1 per million input tokens).
• Conversion Rules: When calling other models, the system will automatically convert and deduct Fusion tokens from your subscription package based on the ratio of that model’s list price to the base model’s price. For example, if a model’s input price is $3 per Mtoken, the number of Fusion tokens consumed when calling that model will be three times that of the base model.
• Consumption Calculation: The token consumption for a single API call is calculated as the sum of input token consumption, cached writes token consumption, cached reads token consumption, and output token consumption, each converted according to the respective model’s deduction factor.

Example: In a reasoning request for "claude-sonnet-4-6"

• Token consumption: Input tokens: 11, Cached writes (5m) tokens: 4629, Output tokens: 2456
• Deduction factor: input: 3, Cached writes (5m): 6.25, output: 25
• Therefore, the total token consumption = 11 × 3 + 4629 × 6.25 + 2456 × 25 ≈ 90,364