Choose a chat model and response speed
Change the model for your next web-chat reply without changing other work.
The web-chat model selector lets you choose among the options enabled for your workspace. Auto uses Artie's current default route; a model name alone is not a guarantee of price, speed or quality.
Available models, thinking levels and speed options depend on the enabled provider configuration. Use the choices actually shown in your chat.
On this page
Choose in the current chat
Open the model selector in Home or the Artisan's Chat composer. Choose Auto or an available model, then send your next request. Changing the choice keeps your draft and affects subsequent replies in that chat; it does not regenerate the previous answer.
Review thinking and speed choices
When the selected model offers a thinking level or Fast/Priority and Slow/Flex options, choose the setting that fits the job. Faster service and more reasoning can have different usage costs. Read the option's explanation rather than assuming every model supports the same settings.
Keep other work separate
A chat selection does not change the workspace default, scheduled task work, helper calls, independent review or speech. Slack has its own selection behavior.
Check the actual reply and result
Use the model information attached to the reply to see what actually answered. Review the output against its sources and your requested outcome. If an option is unavailable or a reply fails, use the current error and recovery path; do not assume the request succeeded merely because a model was selected.
Need help with your workspace?
Contact support