IT Home On November 12, DouBao officially announced that it has completed a product update, adding more available models, and simultaneously introduced a quota adjustment mechanism based on actual consumption.
Previously, DouBao only provided two self-developed models—DouBao 2.1 Lite and DouBao 2.1 Pro—on the model side. After this update, Kimi K3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash became the first third-party models in the selectable list. Users can switch between self-developed and third-party models based on their tasks.
Before that, on October 29, DouBao Work opened the three-party model configuration to both team subscriptions and enterprise subscriptions customers. Enterprises do not need to purchase from the model provider or apply for additional access; the usage is directly deducted from the existing AI points. The update on November 12 will expand this capability to a broader range of users, along with a limited-time discount program for all users.
After introducing multiple models, cost differences become an unavoidable issue. The resource consumption of different models when completing the same task varies. DouBao’s approach is to directly indicate the quota consumption multiplier for each model on the model selection page. The lower the multiplier coefficient, the less quota is consumed to complete the same task.
IT Home note: The quota consumption rate can be understood as the “pricing factor” when different models complete the same task. It does not change the complexity of the task itself, but determines how quickly the available quota in the user account is consumed, allowing users to make a visible trade-off between effectiveness, speed, and cost. The purpose of this design is that users no longer need to guess intuitively which model is “more cost-effective”; the basis for selection is placed directly on the operation interface.
The limited-time discount operates at two levels. First, during the designated discount periods on weekdays and holidays, using the models participating in the activity allows for a discount on quota consumption. Second, for DouBao 2.1 Lite and DeepSeek V4.1 Flash, the quota consumption rate is directly reduced. These two models are designed for high-frequency lightweight tasks, and with the reduced rate, the same tasks consume less quota, making it more durable.
Additionally, DouBao still offers free services for simple, small tasks. The official chart shows that the first item in the model selection list is “Automatic (small, free)”, and the system will automatically match a model based on the task size. When the complexity of the task increases, users can manually switch to DouBao 2.1 Pro or third-party models, and billing will be done at the corresponding rate.
Related Reads: