- Start
- 50 welcome credits for new accounts
- Know the cost
- Exact credit estimate before each run
- Keep the work
- Prompts and results stay in one history
Judge the output, not the promise
One brief can become a campaign, not just a pretty picture
Start with the largest format you need. Reuse the accepted subject, lighting, and copy rules across crops instead of prompting every placement from scratch.



Start with what you already have
A direct route for every image job
01 / New image
Generate from a description
Best when the idea exists but the visual does not.
02 / Existing reference
Change one thing, preserve the rest
Upload the source, name the edit, and list what must stay unchanged.
03 / Finished asset
Cut out, clean up, and upscale
Prepare the accepted result for product, social, or design use.
A usable first draft
Give the model a production brief
Style words alone do not define success. A deliverable, a locked constraint, and a review target make the next iteration obvious.
- 01
Write the deliverable
Say whether this is a product shot, poster, thumbnail, concept frame, icon, or social asset.
- 02
Lock what matters
Quote exact copy and name the subject, crop, lighting, brand details, or reference traits that must stay stable.
- 03
Inspect before you ship
Check text at 100% zoom, then review faces, hands, edges, logos, and small product geometry.
Model choice
Match the model to the review
ByteDance / Seedream 5 Lite
Create an image from a text prompt.
NewByteDance / Seedream 5 Pro
Create a high-quality image from a text prompt.
NewOpenAI / GPT Image 2
Create an image from a text prompt.
HotKrea × Black Forest Labs / FLUX.1 Krea [dev]
A 12B FLUX.1 collaboration tuned for distinctive, natural-looking text-to-image work.
HotBefore you generate
AI image generator FAQ
Can I try the AI image generator for free?+
New accounts receive 50 welcome credits with no credit card required. The generator shows the current credit cost before every run. Additional generations require purchased credits or a plan.
Can I create from text and edit an existing image?+
Yes. Start with a written prompt for a new image, or switch to an edit workflow and upload a reference. For more specialized edits, Voor also has dedicated image-to-image, background removal, translation, restoration, and upscaling tools.
Which AI image model should I choose?+
Use the visible model picker as the current source of truth. GPT Image 2 is a strong starting point for readable copy and structured layouts; other models may be better for fast drafts, reference-led edits, or a particular visual style.
Are generated images private?+
Generations made with welcome credits are public and include a small watermark. After any paid plan, new creations are private by default and downloads are watermark-free.
How do I get a more useful first result?+
Name the deliverable, subject, composition, lighting, and any exact text. Add one constraint about what must stay unchanged. Review spelling, faces, hands, logos, and fine product details before publishing.
Your brief is the starting point
Create the first image, then make it production-ready
Pick a starting brief above, choose the current model that fits the job, and review the exact cost before you run it.


