.png?fit=max&auto=format&n=J3TfNmZhqEoKcaaO&q=85&s=468b5adb026aa33181cc81ab54ab68db)
The three tiers
| Tier | Speed | Intelligence | Credit cost |
|---|---|---|---|
| Lite | Fastest | Good for simple tasks | Lowest |
| Pro | Balanced | Strong across all tasks | Medium |
| Max | Slower | Deepest reasoning | Highest |
When to use each tier
Lite
Use Lite for straightforward requests where speed matters more than depth. Lite responds the fastest and uses the fewest credits.Good for: Quick text edits, simple code fixes, short answers, reformatting content, factual questions.
Pro
Use Pro for most work. It strikes the best balance between quality and cost. If you are unsure which tier to pick, start here.Good for: Code generation, presentations, documents, websites, image and video generation, research.
Max
Use Max when the task requires deep thinking. Max uses the most credits per message but produces the most thorough results.Good for: Complex multi-step projects, detailed analysis, large codebases, nuanced writing, tasks where the quality of decisions matters more than speed.
Chat Mode
In Chat Mode, you select a specific model from a dropdown instead of a tier. Your selection is saved in your browser and persists across sessions. Chat Mode is for conversations only. The agent does not execute tools, generate media, or write code in Chat Mode. See Agent Mode vs Chat Mode vs Plan Mode for the full comparison.What you cannot do
- You cannot use multiple tiers in the same conversation. The tier is set when you send your first message.
- You cannot set a default tier globally. You select it per chat.