The models you can pick, what the effort setting does and which plan includes what.
Use the model button next to the input to pick a model. Models you cannot use on your plan are shown with a badge (Pro or Ultra) and are locked.
| Model | Provider | Good for |
|---|---|---|
| Claude Sonnet 5.5 | Anthropic | Fast and capable for everyday work. The default. |
| Claude Opus 5.5 | Anthropic | The toughest challenges. |
| Gemini 3.6 Flash | Enhanced reasoning and complex work. | |
| Gemini 3.5 Flash | The most efficient for everyday tasks. | |
| Gemini 3.1 Flash Lite | The fastest for quick answers. | |
| Gemini 2.5 Pro | The toughest challenges. |
The list is loaded from the server, so it can change over time. Your last choice is remembered on this device.
| Model | Free | Plus | Extended |
|---|---|---|---|
| Claude Sonnet 5.5 | ✓ | ✓ | ✓ |
| Gemini 3.1 Flash Lite | ✓ | ✓ | ✓ |
| Gemini 3.5 Flash | ✓ | ✓ | ✓ |
| Gemini 3.6 Flash | ✓ | ✓ | |
| Claude Opus 5.5 | ✓ | ||
| Gemini 2.5 Pro | ✓ |
The effort setting (Low, Medium, High) tells the model how much to think before answering.
The highest effort you can use depends on your plan: Low on Free, up to Medium on Plus, up to High on Extended. A higher choice is lowered automatically to what your plan allows.
In Peridot Code and the CLI the same models drive the agent. Both Claude and Gemini models can use the tools (reading, editing, running commands), so pick by speed and depth rather than by ability to use tools. For big refactors or tricky bugs, a larger model with higher effort is usually worth it. For quick edits, a fast model is enough.
AI team bots can each use a different model. A cheap, fast model suits an exploring bot, and a stronger one suits a reviewer. See Creating bots.
Type /model to see the list, or /model <name or number> to change it. See Peridot CLI.