Unlimited Mode lets you create on supported models without consuming tokens. It is designed so you can focus on creating without worrying about usage limits or running out of tokens.
This service is available only on select Plus and Pro plans.
1. What “Unlimited” Means
Within the supported models and resolution ranges, you can generate as much as you want without overall token consumption.
* No token deduction: Tokens are not consumed for overall usage.
* No generation cap: There is no limit on total number of generations.
* Ongoing access: You can continue using it as long as the models included in your plan remain available.
* Core value: Unlimited means access, not speed.
Unlimited mode does not limit how much you can generate, but processing speed may be adjusted for system stability. Unlimited mode supports a certain number of boost generations per plan so you can generate at maximum speed.
Boost generation:
* Priority: Jobs get priority in the queue.
* Plan-specific boost availability: Not guaranteed on all plans or models; availability varies by plan. Higher-priced plans support more boost generations.
* When exhausted: Even after using all boosts, generation does not stop and unlimited generation continues. However, queue priority is lower, and generation may be delayed when the queue is busy.
Unlimited mode wait times
In unlimited mode, wait times can occur depending on the number of pending jobs from users on the same generation mode among AI modes that support unlimited generation. Depending on this situation, generation time may increase. It can take as little as 3 minutes or as long as 2–5 hours, and in some cases more than 24 hours depending on pending job volume. Even if users are dissatisfied with unlimited mode wait times, the service provider is not responsible. If unlimited wait times are long, we recommend canceling unlimited generation and generating using tokens instead.
Automatic cancellation during long unlimited waits
In unlimited mode, wait times can occur depending on the number of pending jobs from users on the same generation mode. If this wait is expected to be very long, generation may be automatically canceled. No tokens are deducted when generation is canceled, but users must try generating again if canceled. If unlimited wait times are long or generation is automatically canceled for this reason, we recommend generating using tokens instead.
4. Usage reset and verification
* Reset cycle: Boost usage resets every 30 days based on your subscription billing date.
5. Responsible use
Unlimited mode is designed to support personal, human creativity. To maintain stable service, the following are prohibited:
* Automation and scraping: Automated generation or data collection using bots or similar bot-like methods.
* Account sharing and resale: Sharing one account among multiple people or commercially reselling generated content.
* Pattern monitoring: Usage patterns may be reviewed to ensure fair use.
If violations of the above are found, account restrictions may be applied without warning. 