How to Use Nebula
From zero to a perfect AI generation prompt in under a minute. Here's everything you need to know.
Describe your vision (or upload an image)
Open the Nebula Studio and you'll see the main prompt input box — a clean, minimal text area.
Type any description of what you want to generate. Be as detailed or as vague as you like. For example: "A lone astronaut on a neon-lit alien world" or simply "Cozy coffee shop at dusk".
Alternatively, click the image widget in the top-right corner of the prompt box to upload a reference image. You can also drag and drop an image directly onto it.
💡 Tip: You can combine both — upload an image AND add text context for more precise results.
Upload an image (optional)
The compact image widget lives in the top-right corner of the prompt box. It's small by design — it expands with a satisfying spring bounce when you hover or drag an image over it.
Supported formats: JPG, PNG, WebP, GIF (max 50 MB).
Once an image is attached, you'll see a thumbnail preview inside the widget. Click the × to remove it.
When an image is provided, Nebula uses AI vision to analyze composition, lighting, color grading, style references, and technical parameters.
💡 Tip: Screenshot a design, artwork, or photo from anywhere and drop it in — Nebula will reverse-engineer the prompt.
Generate your Master Prompt
Click the Generate button (or press Enter) to start the AI analysis.
A subtle loading indicator appears — the generation typically takes 3–8 seconds.
Nebula's AI engine (powered by Gemma 4 (31B), via OpenRouter) analyzes your input using a specialized reverse prompt engineering system. It extracts: subject, composition, lighting, color palette, artistic style, camera settings, atmosphere, and quality modifiers.
The result is a 150–400 word Master Prompt — dense, comma-separated, optimized for top generative models.
💡 Tip: The prompt is specifically tuned for Midjourney, DALL-E 3, Stable Diffusion XL, and Flux.
Copy and use your prompt
Once generated, the Master Prompt appears below the input box.
Click the Copy button in the top-right of the output card to copy the full prompt to your clipboard.
Paste it directly into Midjourney (/imagine), DALL-E 3, Stable Diffusion, or any other tool.
Your prompt is also automatically saved to your Recent Explorations history below the prompt box.
💡 Tip: Prompts are saved locally in your browser. Sign in to sync them to the cloud and access from any device.
Browse your history
Every prompt you generate is saved to your Recent Explorations list, visible directly below the prompt box.
Click any history item to restore its generated prompt in the output area.
Signed-in users get cloud history synced across devices.
Contribute to the Community
Love a prompt you generated? Share it with the Neptune community.
Click 'Contribute to Community' below the output — the prompt will be published to the community feed at /nebula/community.
Other creators can discover, use, and be inspired by your prompt.
You must be signed in to contribute. The Sign In button is in the top navigation bar.