Text to image
Describe the subject, scene, composition, lighting, and style to create a new image.
AI image generation workflow
Generate AI images with Orqra
Organize text-to-image and reference-image creation as a reusable local-first workflow. Keep subjects, scenes, styles, prompts, and generated results with your project while choosing the model for each task.
Three ways to create an AI image
Single reference
Use one source image to guide a new result when the selected model supports it.
Multiple references
Combine several reference images for visual direction or style with models verified for that input.
Verified AI image models
Every model below has passed Orqra capability verification. The labels show the specific text or reference-image inputs verified for each model.
- Nano Banana ProText to imageSingle referenceMultiple references
- Nano Banana 2Text to imageSingle referenceMultiple references
- Nano Banana 2 LiteText to imageSingle referenceMultiple references
- GPT Image 2.5 SunburstText to imageSingle referenceMultiple references
- GPT Image 2.5 FlareText to imageSingle referenceMultiple references
- GPT Image 2Text to imageSingle referenceMultiple references
- FLUX.2 MaxText to imageSingle referenceMultiple references
- FLUX.2 ProText to imageSingle referenceMultiple references
- FLUX.2 FlexText to imageSingle referenceMultiple references
- Recraft V4.1Text to imageSingle reference
- Recraft V4 StylesMultiple references
From an idea to a reusable image workflow
Prepare the content
Add the subject, scene, style, and any reference images the task needs.
Structure the prompt
Keep reusable creative content separate from the instruction for this image.
Choose a model and run
Select a model that matches the input mode, review its settings, then save and reuse the result.
Write clearer prompts with Prompt Fabric
Prompt Fabric keeps reusable creative content separate from task instructions. Save recurring descriptions and directions as templates, insert them when needed, and adjust each copy for the shot. It organises the request; it does not guarantee model output quality.
Learn Prompt FabricAI image generation questions
How do I generate an AI image in Orqra?
Connect OpenRouter, add text or supported image inputs, choose a verified image model, write the prompt, and run the node.
Which AI image models does Orqra support?
The current verified list includes GPT Image 2.5, Nano Banana, FLUX.2, and Recraft models. Text and reference-image modes differ by model.
Do I need cloud storage to generate images?
No. Standard image generation does not require S3-compatible storage. Storage is optional for workflows where a provider needs a public media URL, such as some video tasks.