Select models in Qoder CN, understand Credits usage reference factors, and configure model parameters.
Qoder CN provides flexible model selection to help you balance task quality, response speed, and Credits usage. Use Auto for automatic routing or choose a specific available model.
Start with Auto for general tasks. Choose a specific model when you have a model preference or need a particular capability.
Select a model directly when you have a specific preference or task requirement.
Click Model Management at the bottom of the selector and configure parameters in the Default category. Supported parameters and available values vary by model.
The context window determines how much context the model can use when processing a request. Adjust the slider in the model's Context window column to select a supported size.
Thinking intensity adjusts the model's reasoning depth. Open the Thinking intensity dropdown for a supported model. For example, Qwen3.8-Max offers thinking off, low, medium, and extra high.
Models showing an unsupported or fixed thinking status do not offer the same adjustable options. Use the controls available for the selected model.
Click Save settings to apply changes across tasks. A response already in progress keeps its current settings; new settings apply from the next turn. Click Cancel to discard changes.
In the Default category of Model Management, use Display status toggles to show or hide models in the selector. To configure your own API key, switch to Custom. See Custom Models for instructions.
Yes. Select another model in the input box. The new selection applies to subsequent messages.
Auto selects a suitable model according to its routing strategy. Select a specific model when you know which model you want to use.
Available models change, and promotions can affect displayed factors. Refer to the models and factors currently shown in the client.
First check its display toggle in Model Management. If it is also absent from that list, refer to the models currently available to your account and client.
Credits usage depends on the tokens used by each request and the selected model. Reference factors compare estimated usage for similar tasks; they are not a fixed fee per request. See Credits.
You can also connect your own models with an API key. See Custom Models.
Auto routing
| Model | Description | Use case | Credits usage reference factor |
|---|---|---|---|
| Auto | Automatically selects a suitable model to balance performance and cost | Everyday tasks without a specific model requirement | 0.5× |
Select a specific model
Select a model directly when you have a specific preference or task requirement.
| Model | Description | Credits usage reference factor |
|---|---|---|
| Qwen3.8-Max | Qwen foundation model for software engineering, professional work, and deep reasoning | 0.5× |
| Qwen3.8-Flash | Multimodal MoE model balancing capability, latency, and cost | 0.1× |
| Qwen3.7-Max | Agentic model for long-running, complex tasks | 0.5× |
| Qwen3.7-Plus | Reasoning and agentic capabilities for coding and complex problem solving | 0.1× |
| DeepSeek-V4-Pro | Model for agentic tasks, knowledge, and reasoning | 0.5× |
| DeepSeek-Flash | Model supporting agentic tasks, knowledge, and reasoning | 0.1× |
| GLM-5.3 | Model for complex systems engineering and long-running tasks | 0.8× |
| GLM-5.3-Flash | Multimodal model for image and video understanding, research, and document creation | 0.1× |
| Kimi-K3 | Model for software engineering, knowledge work, and deep reasoning | 1.4× |
| Kimi-K2.8-Preview | Model for long-context coding, instruction following, and extended tasks | 0.8× |
| MiniMax-M2.7 | Native multimodal perception and frontier coding capabilities, suited for high-intensity workflows | 0.2× |
Refer to the client for available models and current factors. During a promotion, the selector may show regular and discounted factors alongside an off-peak discount notice. Check the applicable promotion rules for eligibility and duration.
Configure model parameters
Click Model Management at the bottom of the selector and configure parameters in the Default category. Supported parameters and available values vary by model.
Context window
The context window determines how much context the model can use when processing a request. Adjust the slider in the model's Context window column to select a supported size.
Thinking intensity
Thinking intensity adjusts the model's reasoning depth. Open the Thinking intensity dropdown for a supported model. For example, Qwen3.8-Max offers thinking off, low, medium, and extra high.
Models showing an unsupported or fixed thinking status do not offer the same adjustable options. Use the controls available for the selected model.
Save settings
Click Save settings to apply changes across tasks. A response already in progress keeps its current settings; new settings apply from the next turn. Click Cancel to discard changes.
Open model settings
In the Default category of Model Management, use Display status toggles to show or hide models in the selector. To configure your own API key, switch to Custom. See Custom Models for instructions.
Switch models
1
Open the selector
Click the model selector at the bottom of the input box in a new or existing task.
2
Select a model
Choose Auto or an available model.
3
Continue the task
Check the selected model and reference factor before sending. The selection applies to subsequent messages.