---
title: Text Generation
slug: text-generation
createdAt: 2026-06-22T07:17:36.333Z
---

:::hint{type="success"}
Text generation workflows use large language models (LLMs) to generate, rewrite, summarize, and analyze text from prompts. Use them for content creation, coding assistance, brainstorming, research, documentation, and Q\&A.
:::

## Overview

Text generation workflows use LLMs to produce and transform text from your prompts. This example uses the **Qwen3-5-9B** model to generate text from a user prompt, with optional image input, so the model can understand both text and visual content.

:::hint{type="success"}
**MULTIMODAL: **(note style) Because this model accepts images as well as text, you can ask questions *about *an image, extract information from it, or generate descriptions, not just work from text alone.
:::

## What you'll learn

- Generate text from prompts
- Use LLMs for content creation
- Generate code and technical content
- Analyze images alongside text prompts
- Adjust generation settings for different outputs

## Example Workflow

:::CtaButton{label="Qwen3-5-9B: Open-Source LLM & VLM →" externalHref="https://www.floyo.ai/workflows/qwen-3-5-9b-for-open-source-llm-and-767bx08eghx8" openInNewTab="true"}

:::

1. **Enter your prompt:** In the **String (multiline)** node, type the instruction or question you want the model to respond to. Examples: 

- *Write a product description for a smartwatch*
- *Summarize this article in three paragraphs*
- *Create a Java calculator application*
- *Explain machine learning in simple terms*

![](https://api.archbee.com/api/optimize/1hT0XoAHMhU4co8ZZjygd/mqR2BzU1R5mg1Jgjafmnv_image.png)

2. **Configure generation settings:** The **TextGenerate** node controlshow the model responds (see the settings table below). For most use cases, the defaults work well without changes.

::Image[**(Optional) Add an image:** Connect an image to the workflow if you want the model to analyze visual input alongside your prompt.]{src="https://api.archbee.com/api/optimize/1hT0XoAHMhU4co8ZZjygd/HGWY63KQIgSJnuDLx2CKW_image.png" size="50" isUploading="false" width="366" height="468" darkWidth="366" darkHeight="468" showCaption="false"}

3. **Run the workflow:** Click **Run**. The model processes your prompt (and image, if provided) and generates a response.
4. **Review the output:** The response appears in the **Preview as Text** node. Review it, copy the text, and use it in your projects.