LLM Tool

Need help creating a prompt, refining an idea, or turning some rough context into something ready to use? The LLM tool lets you chat with AI models directly inside Fluidfit ...or write a prompt right on the node — and turn the result into a reusable prompt for the rest of your workflow.
​

No need to jump between tools, copy prompts around, or start from scratch every time.
​

Add the LLM tool from the Advanced category.
​
You can now also type a prompt directly on the node and click Generate to run it right away — no need to open chat at all. Or leave the prompt empty and click Open chat for a back-and-forth conversation instead.

Choose the AI model you want to work with, then use the conversation to:

The LLM tool accepts optional Images, Videos, and Text inputs.

Connect the content you want the AI to work with and reference it in your conversation using #.

For example:

This makes it easy to turn your existing content into something ready for the next step.

Click the edit icon next to the prompt field to open the Prompt Editor — a larger writing space with undo, redo, and copy controls, and the same # reference support. Save to bring your prompt back into the node, or Cancel to discard changes.

Different AI models excel at different types of work. Select the model that best suits your use case before starting a conversation.

If you'd like to compare responses from multiple models, simply create another LLM node and start a separate chat.

Using dedicated LLM nodes for different tasks keeps your workflow organized and makes it easy to reuse prompts later.

​

Once a conversation has been created, the node displays your latest prompt for quick reference. Click Generate to run the LLM with that prompt and pass the latest response to the connected Text node.

Once the LLM generates a response, it is passed to the connected Text node.
​

The Text output is editable, so you can even review the generated prompt, make your own changes, and fine-tune it before sending it further down the flow.
​

From there, you can use the Text output as a prompt in supported tools such as Generate Image, Generate Video, and AI Edit.
​

Simply connect the Text node to the tool and turn on Use this prompt.

This gives you a simple workflow:

Context → LLM → Edit text if needed → Generate Image, Generate Video, AI Edit, or another supported tool​

You can create multiple LLM nodes in the same flow.

This is useful when you have different prompt requirements — for example, one LLM node for creating an image prompt and another for creating a video prompt.

Keeping them separate makes your flows easier to understand, reuse, and update.

Each LLM node has its own conversation history.

You can:

This makes it easy to iterate without losing your earlier ideas.

​

The LLM panel also gives you a quick view of your model usage, including:

Usage is tracked at the account level for each model, rather than separately for each chat session.