Hello, this is Plaitoon.

Starting July 9, the Output Length Adjustment feature will be implemented.

The Output Length Adjustment feature allows you to set response lengths more flexibly according to the work and the conversational context. You can choose the dialogue experience you want, ranging from faster-paced exchanges to more enriched responses.

Effective Date
July 9, 2026, at 19:00

Basic Allowance
All models include a base of 1,200 tokens.

Event Information
To help you experience this feature more conveniently during the initial rollout, we are providing an additional 300 tokens for free until July 31.

For example, even if you set the maximum output for Gemini 2.5 Pro to 1,500 tokens, gems are not deducted based on 1,500 tokens every time. If the actual response only uses 1,300 tokens, only the 100 tokens exceeding the 1,200-token basic allowance will result in an additional deduction. In this case, for Gemini 2.5 Pro, the deduction would be Base 72 Gems + Excess 6 Gems = Total 78 Gems. The maximum output is simply the upper limit for how long a response can be; actual deductions are calculated based on the tokens used in the generated response.

---

Deduction Policy Guide

The Output Length Adjustment feature does not always deduct the maximum amount just because you have set a high maximum output.

Actual deductions are calculated based on the tokens actually used, not the maximum output length set.

• If the actual usage is 1,200 tokens or less, only the base gems for each model will be deducted.
• During the event period, if the actual usage is 1,500 tokens or less, no additional gems will be deducted.
• Additional gems are deducted only when usage exceeds the event allowance, and only for the excess tokens.
• Excess usage is calculated in 100-token units.

For example, even if you set Gemini 2.5 Pro to a maximum of 1,500 tokens, if the actual response only uses 1,200 tokens, only the base 72 gems will be deducted. The maximum output is simply the upper limit for how long a response can be; actual deductions are calculated based on the tokens used in the generated response.

---

Base and Excess Deduction Standards by Model

If you have any inconveniences or feedback regarding output length, deduction methods, or conversation length while using the service, please feel free to contact Customer Support.

We will continue to improve based on your feedback to create a better conversational experience.

Thank you.
Best regards, The Plaitoon Team