Accessing the model selector
Open the AI assistant and look for the model selector at the top of the chat panel, near the conversation title. It displays the name of the currently active model. Click it to open a dropdown of available models.
Available models
Userorbit offers several AI models, each with different strengths. The exact models available depend on your plan, but typical options include:
- Fast models — Optimized for speed and lower cost. Responses arrive quickly and work well for straightforward tasks like categorizing feedback, generating short responses, or answering factual questions about your data.
- Balanced models — A middle ground between speed and reasoning depth. Suitable for most everyday tasks including drafting articles, summarizing feedback trends, and composing announcements.
- Advanced models — The most capable option, designed for tasks that require nuanced reasoning, long-form content, or complex multi-step analysis. Use these when quality matters more than speed.
Choosing the right model
Consider the task at hand:
- Quick answers and simple tasks — Use a fast model. If you are asking the agent to categorize a piece of feedback or draft a one-line acknowledgment, a fast model handles it efficiently.
- Content drafting — Use a balanced or advanced model. Drafting help center articles, changelog entries, or customer announcements benefits from the deeper reasoning of these models.
- Analysis and strategy — Use an advanced model. When you need the agent to analyze feedback trends, compare feature requests, or generate a detailed summary, the advanced model produces more thorough results.
Changing the model mid-conversation
You can switch models at any point during a conversation. The new model takes effect for subsequent messages only — previous responses are not regenerated. This is useful when you start with a fast model for brainstorming and then switch to an advanced model for the final draft.
Model and cost
Each model consumes AI credits at a different rate. Fast models use fewer credits per message, while advanced models use more. Your current credit usage is visible in your workspace settings under the billing section. If you are on a limited plan, using fast models for routine tasks helps you stay within your allocation.
Best practices
- Default to balanced. The balanced model handles most tasks well and is a safe starting point.
- Upgrade for polish. When a draft needs to be customer-facing or published publicly, switch to the advanced model for the final pass.
- Downgrade for volume. If you are processing a large batch of feedback responses, use a fast model to conserve credits.
- Experiment. Try the same prompt on different models to see how the output varies. This helps you develop intuition for when to use each one.