Skip to main content
NodePad does not bind a thread to a single model. Every message you send has its own model selector, so you can change models at any point without starting a new conversation or losing your context. This makes it practical to use the right model for each part of your work rather than settling for one model across everything.

​
How to select a model per message

1

Open the model selector

In the message input area at the bottom of any thread node, click the model name shown to the left of the send button. A dropdown appears listing all available models.
2

Choose your model

Select the model you want to use for this message. The selector updates to show your choice.
3

Send your message

Send the message as normal. The response is generated by the model you selected. The next message defaults back to your canvas-level default model, or you can change it again before sending.
You can set a default model for the entire canvas from the canvas settings panel. Per-message selection overrides that default for individual messages only.

​
Running models in parallel across threads

Because every thread on the canvas is independent, you can assign different models to different threads and run them at the same time. This is one of NodePad’s core advantages over single-thread AI tools. To compare models on the same question:
1

Create a second thread

Fork any message or open a new blank node on the canvas. You now have two separate threads.
2

Set a different model in each thread

In the first thread, select Claude. In the second, select GPT-5 or Gemini. Each thread’s model selector is independent.
3

Send the same prompt to both

Type the same question in each thread and send. Both responses generate simultaneously, and you can read them side by side on the canvas.

​
Use cases for per-message and parallel selection

You’re deep into a research thread with Claude and it gives you an answer you’re not sure about. Rather than starting over, fork the message and route the fork to GPT-5. You get a second perspective without losing the conversation history that led you there.
Assign one model to a thread covering technical sources and another to a thread covering qualitative or creative angles. Run both simultaneously and synthesize the results on the canvas, rather than running the same model twice and hoping it gives you diverse output.
Writers can send the same brief to Claude, GPT-5, and Gemini in three parallel threads. Each model has a different default tone and style, and seeing all three drafts at once makes it easy to pick a direction or blend elements from each.
Use Gemini 2.5 Pro for a long-document analysis thread while running a coding question through Claude Opus in a separate thread. You’re not picking one model for the whole session — you’re picking the best model for each task.
All threads on your canvas share the same session. Switching models does not affect your billing tier or require separate logins for each provider — NodePad handles authentication centrally.