Switch Models in the Middle of a Project in Freebuff Desktop
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
Switch Models in the Middle of a Project in Freebuff Desktop
Different tasks suit different models. This example starts a project on one model in Freebuff Desktop, switches in the same chat and carries on.
Quick answers
Can I switch models in Freebuff Desktop? Yes. Open the model selector at the bottom of the input box, pick another model and send your next message.
Does switching start a new chat? No. The chat and the files it created carry over, and the next message uses the new model.
What does the picker offer? A speed slider with Fast, Balanced and Powerful, models grouped under those headings, and a reasoning level with Low, High and Max.
Where do I see how a model handles data? Each row can carry a note such as May use data for AI training. Pick the model that fits your project.
Can each chat use a different model? Yes. Each chat has its own selector. When two chats run side by side, each shows its own model.
Open the picker
Open a folder as a project in Freebuff Desktop and start a new chat. The input box has a selector showing the current setting, Balanced here. Open it:
Fast Balanced Powerful Optimized DeepSeek V4.1 Flash 10/17 Recommended Muse Spark 1.3 GLM 5.3 Flash GPT-6 Luna 3 more Powerful DeepSeek V4.1 Flash Fast 10/17 Claude Haiku 5.5 (100k context) May use data for AI training Reasoning: Low High Max
Model names and groupings change often. The note about training data appears under the models it applies to.
Work on the first model
Pick any model and send:
Write fizzbuzz.py that prints FizzBuzz for 1 to 15 and run it.
Freebuff wrote the file, ran it and printed the 15 lines.
Switch and continue
Open the selector again, choose a different model and send a follow-up in the same chat:
Add a command line argument to fizzbuzz.py for the upper limit and run it with 20.
Freebuff reported that the script now takes an optional limit, defaulting to 15, and ran it with 20.
Verify the result
python3 fizzbuzz.py 20 | tail -4
17 Fizz 19 Buzz
The file written on the first model was edited by the second, in the same chat.
Troubleshooting
Which model answered. Check the selector under the input box before you send, since it shows the model for your next message.
A model is missing. The list changes over time. Choose another, or open the section that says more to see the rest.
Private code. Choose a model whose row has no data note.
Tested in Freebuff Desktop 0.0.184 on 2026-10-30.