> For the complete documentation index, see [llms.txt](https://docs.diaflow.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.diaflow.io/workflow-builder/nodes/private-ai-llm-models/llama.md). # Llama ### 1. The "Ready-to-Use" AI (Overview) Zero setup, zero API keys. Just drag, drop, and start prompting using Diaflow's built-in AI credits. Llama’s superpower is flexible reasoning with optional vision on selected models. It is a strong fit when you want one AI node for text-heavy work, conversation-style workflows, and image analysis in the same builder. ### 2. How to Add the Node Click the **+** button on your canvas, search for **Llama**, and click to add it to your workflow. > 🖼️ **\[PLACEHOLDER: I will insert a GIF here showing how to search and add the node onto the canvas]** ### 3. What can you automate? (The "Wow" Use Cases) #### 1. **The Massive Document Analyzer:** Send long reports, contracts, proposals, or knowledge-base text into Llama and turn them into fast summaries, risks, and recommended next steps. #### 2. **The Multimodal Scanner:** Pass in an image link for a receipt, invoice, product photo, or screenshot and let Llama pull out the details your team needs. #### 3. **The Content/Data Factory:** Turn messy notes, form submissions, or research text into structured business content, polished copy, or cleaner downstream data. ### 4. Step-by-Step: Configuration & Smart Data Injection #### 1. **Choose your Brain (Model):** Select your model from the dropdown. Use smaller models like **1B**, **3B**, or **8B** for faster, lighter tasks. Use larger models like **70B**, **90B**, **Maverick**, or **Scout** for deeper reasoning and more demanding work. For image tasks, choose a vision-capable model such as **Llama 4 Maverick**, **Llama 4 Scout**, or a **Llama 3.2 Vision** model. #### 2. **Write the Prompt & Inject Data:** The easiest way to add live workflow data is to type `@` on your keyboard. Diaflow opens a visual dropdown so you can pick values from earlier steps, like a customer name, a form response, or text from another node. Typing `{{` also works perfectly if you prefer the classic bracket syntax. #### 3. **Attach Media (If applicable):** If you want Llama to analyze an image, add the image link in the image field. This is useful for receipts, screenshots, product images, and scanned visuals. > 🛑 **\[PLACEHOLDER: I will insert my real UI screenshot showing the Prompt and File setup here]** ### 5. The Output: Seeing Your Results This node returns the generated result so you can pass it straight into the next step. Use `{{node_name.data}}` in your next step, like Gmail or Google Sheets, to pass the generated text or media forward. You will also see `input_token` and `output_token`. Since Diaflow handles the billing, these tokens act as a meter so you can track how many credits your workflow is consuming. The node also returns the model used and the final resolved prompt, which helps you review what ran. > 🛑 **\[PLACEHOLDER: I will insert my real Execution Result snippet here]** ### 6. Golden Rules & Guardrails (CRITICAL PRO-TIPS) #### **File/Image Limits:** Only image input is supported in this node for multimodal analysis. Keep these rules in mind: * only the **first image** in a request is processed * additional images are ignored * supported image formats are `jpeg`, `png`, `gif`, and `webp` * if the image is too large, the run can fail and you may need a smaller image If image analysis matters, test with one clear image first. #### **The System Prompt Trap:** The separate system instruction field is currently not applied in this node. For guaranteed results with advanced models, put your Persona/System instructions directly into the main Prompt box. #### **Auto-Streaming:** The AI will think quietly and return the full complete answer at once. #### More pro-tips * Not every Llama model supports images. Choose a vision-capable model for image tasks. * Conversation memory is available in this node, so Llama can keep context across turns when memory is enabled. * Caching is available if you repeat the same request often and want faster reuse. * If you see **Invalid Model**, recheck the model you selected. * If you see **Image is too large**, reduce the image size and try again. * Use the active **Llama** node for new workflows, not the legacy version. ### Need help? * Learn the basics in [How a node works](/getting-started/lets-start-with-the-basics/how-a-node-works.md) * Browse available nodes in [Component List](/workflow-builder/component-list.md) * Build the full flow in [Create a workflow](/workflow-builder/create-a-workflow.md) --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.diaflow.io/workflow-builder/nodes/private-ai-llm-models/llama.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.