Creator
$19/month
Billed monthly
For solo creators posting every week
500 credits every month
- Up to 500 GPT Image 2.5 Sunburst images
- Up to 50 AI videos
- Up to 1.5 hours of AI voiceover
- Roll over up to 1,000 credits
AI image generator Type it. See it.
Describe a picture or add your own photos. Get photos, illustrations, posters, product shots and edits from 29 image models in one place.
Generate an image
Plans from $19 a month. Up to 4K, four images per run.
The prompt, run on GPT Image 2.5 Sunburst
A 1960s style travel poster for Lisbon. Large title at the top reading exactly "LISBOA". A yellow tram climbing a steep street with terracotta roofs and the river behind. Bottom line in smaller letters reading exactly "Linha 28 · Every 10 minutes". Flat screen-print colors, visible paper grain.
Which AI image generator is best?The one built for your job
Not sure? Start with GPT Image 2.5 Sunburst. It is the default when you open the image tool. For these seven jobs, start here instead.
01
Words on a poster, label or menu
GPT Image 2.5 Sunburst
Built for text and designed graphics: posters, ads and thumbnails. Put the exact words in quotes, like "LISBOA".
GPT Image 2.5 Sunburst 02
A photo that looks real
Nano Banana Pro
Photoreal scenes with precise materials and light. Name the lens and the light the way a photographer would.
Nano Banana Pro 
Nano Banana Pro 03
Your product in a new scene
GPT Image 2.5 Sunburst
Add a phone photo of the product as a reference and describe the scene. Ask it to keep the shape, color and label exactly.
Reference 
GPT Image 2.5 Sunburst 04
Illustration or a cartoon character
Nano Banana 2.1
Illustration and cartoon characters, with the Nano Banana look at a lower price. Name the medium: gouache, ink, watercolor.
Nano Banana 2.1 05
The same person in new places
Seedream 5.0 Pro
Edits real photos and keeps the person's identity. Add one photo and say what must stay the same.
Reference 
Seedream 5.0 Pro 
Seedream 5.0 Pro 06
One image from several references
Nano Banana Pro
Combines up to 14 reference images in one picture. Say which image each part comes from.
Reference 
Reference 
Nano Banana Pro 07
A change to an image you have
Image editor
Open the image, type the one change, and say that everything else stays the same.
Reference 
Image editor
Same prompt, different modelsSee how they differ
Each model ran the prompt once. No retries, no edits.
Editorial portrait of a ceramicist in her fifties with grey curls and clay-dusted hands, holding a freshly thrown bowl in her studio. Shot on an 85mm lens at f/2. Soft north light from a large window on the right. Natural skin texture, linen apron, shelves of drying pots behind her.
Four models, four different women. GPT Image 2.5 Sunburst also added two studio signs the prompt never asked for.
A 1960s style travel poster for Lisbon. Large title at the top reading exactly "LISBOA". A yellow tram climbing a steep street with terracotta roofs and the river behind. Bottom line in smaller letters reading exactly "Linha 28 · Every 10 minutes". Flat screen-print colors, visible paper grain.
All three spelled both lines right on the first run. They differ in palette, framing and how much of the city they draw.
How to use an AI image generatorGrow the prompt
- 1
Describe the image
Write what you want to see. Add photos as references if the picture needs your product, your face or your style.
- 2
Pick a model and a size
Keep the default, GPT Image 2.5 Sunburst, or pick one from the list below. Choose the shape, the resolution and up to four images per run.
- 3
Generate, then edit
Keep the best one. Open it in the image editor to change one part, or turn it into a video.
Four runs on GPT Image 2.5 Sunburst. From step 2 on, each run adds one part to the prompt before it.
Prompt at step 4
A flower stall on a narrow city street at dawn, buckets of tulips, ranunculus and eucalyptus, an older man in a green canvas apron wrapping a bouquet in brown paper. Low golden sunlight from the end of the street, long shadows on wet cobblestones. Shot on a 35mm lens at eye level, shallow depth of field, natural color. A small hand-painted wooden sign on the stall reads "Tulips 3 for 10".
AI image generator from an imageChange one thing, keep the rest
Open any image in the image editor and type the change. Drag the slider to compare.
The edit prompt
Make it night, with a full moon over the sea and warm lamps glowing along the pier. Keep everything else the same.
Every image model in Sceneform29 image models, one tool
The image models out of Sceneform's 90+, which also cover video and audio. Pick one in the image tool's model menu. References are the most photos one run can take; none means text only. Largest is the top resolution you can pick; one size means the model has no resolution setting.
| Model | Pick it for | References | Largest |
|---|---|---|---|
| Text, design and products | |||
| GPT Image 2.5 SunburstText, posters, ads, product photos, precise edits. The default. | Text, posters, ads, product photos, precise edits. The default. | 16 | 4K |
| GPT Image 2.5 FlareA second GPT Image 2.5 look for text, design and edits | A second GPT Image 2.5 look for text, design and edits | 16 | 4K |
| GPT Image 2Text and design work, when you want the older look | Text and design work, when you want the older look | 16 | 4K |
| Ideogram 4.0Text-heavy graphics, posters and logos at a low price | Text-heavy graphics, posters and logos at a low price | None | One size |
| Qwen Image 3.0Text in Chinese and English, graphic layouts | Text in Chinese and English, graphic layouts | 3 | 2K |
| Recraft V4.1Logos, icons and flat illustration in a set palette | Logos, icons and flat illustration in a set palette | None | One size |
| Recraft V4.1 UtilityClean, front-facing product shots and mockups | Clean, front-facing product shots and mockups | None | One size |
| Photos and people | |||
| Nano Banana ProPhotoreal scenes, many references, infographics with web facts | Photoreal scenes, many references, infographics with web facts | 14 | 4K |
| FLUX.3Photographic realism with lens and light control | Photographic realism with lens and light control | 10 | 4K |
| FLUX.2 MaxThe highest FLUX.2 quality, edits from many references | The highest FLUX.2 quality, edits from many references | 10 | One size |
| FLUX.2 ProPhotographic images at a mid price | Photographic images at a mid price | 10 | One size |
| FLUX.2 FlexControl over sampling steps and guidance | Control over sampling steps and guidance | 10 | One size |
| Seedream 5.0 ProEdits that keep a person's identity, many references | Edits that keep a person's identity, many references | 10 | 2K |
| Seedream 5.0 Lite2K to 3K stills at a low price | 2K to 3K stills at a low price | 10 | 3K |
| Soul 2.0Realistic people and candid phone-photo looks, no text | Realistic people and candid phone-photo looks, no text | None | 1080p |
| Kling 3.0 OmniConsistent characters to animate later with Kling video | Consistent characters to animate later with Kling video | 4 | 4K |
| Gen-4 ImageCinematic stills to animate with Runway video | Cinematic stills to animate with Runway video | 3 | 1080p |
| Illustration and style | |||
| Nano Banana 2.1Illustration, cartoon characters, edits from many references | Illustration, cartoon characters, edits from many references | 14 | 4K |
| Grok Imagine 2.0Bold, expressive images and very wide shapes such as 2:1 | Bold, expressive images and very wide shapes such as 2:1 | 5 | 2K |
| Hunyuan Image V3Complex scenes with many described elements | Complex scenes with many described elements | None | One size |
| Fast and cheap drafts | |||
| Z Image TurboThe fastest, cheapest drafts | The fastest, cheapest drafts | None | One size |
| Nano Banana 2 LiteMany variations at low cost | Many variations at low cost | 14 | One size |
| Seedream 5.0 FlashFast, cheap Seedream drafts and edits | Fast, cheap Seedream drafts and edits | 10 | 2K |
| FLUX.2 KleinFast, cheap drafts | Fast, cheap drafts | 4 | One size |
| FLUX.2Cheap photographic drafts | Cheap photographic drafts | 3 | One size |
| Muse ImageCheap everyday images, up to 10 per run | Cheap everyday images, up to 10 per run | None | One size |
| Wan 2.7Low-cost images and edits | Low-cost images and edits | 4 | One size |
| Image-01Very cheap drafts | Very cheap drafts | None | One size |
| Gen-4 Image TurboCheaper cinematic stills | Cheaper cinematic stills | 3 | 1080p |
Simple pricing that scales with you.
Pick the plan for how much you post. Every model on every plan. Unused credits roll over. Cancel anytime.
Pro
Most popular$49/month
Billed monthly
For creators posting daily across channels
1,300 credits every month
- Up to 1,300 GPT Image 2.5 Sunburst images
- Up to 130 AI videos
- Up to 4 hours of AI voiceover
- Roll over up to 2,600 credits
Studio
$99/month
Billed monthly
For brands and production teams
2,700 credits every month
- Up to 2,700 GPT Image 2.5 Sunburst images
- Up to 270 AI videos
- Up to 9 hours of AI voiceover
- Roll over up to 5,400 credits
Scale
$199/month
Billed monthly
For agencies running always-on content
5,500 credits every month
- Up to 5,500 GPT Image 2.5 Sunburst images
- Up to 550 AI videos
- Up to 18 hours of AI voiceover
- Roll over up to 11,000 credits
Every plan includesEvery model, 90+ and countingReusable AI personasSceneform Studio editorCommercial usage rights
Want checkout, comparison tables, or a custom volume? See the full pricing page
Before you generateCommon questions
An AI image generator is a model that turns a written description into a new picture. Many also take your own photos as references, so the result can show your product, your face or your style. Sceneform puts 29 image models in one tool, out of 90+ models for images, video and audio.
Plans start at $19 a month.
It depends on the model and the size. A GPT Image 2.5 Sunburst image costs about 6 credits at its default 2K, high quality, and 1 credit at 1K, medium. Nano Banana 2.1 costs 2 credits at 1K and Nano Banana Pro 8 at 2K. The $19 Creator plan has 500 credits a month, up to 500 Sunburst images at 1K. The image tool shows the cost before you generate.
The best one depends on the job. GPT Image 2.5 Sunburst is the default and the safest start for text, posters and product photos. Nano Banana Pro is strong on photoreal scenes and many references. Nano Banana 2.1 suits illustration. Seedream 5.0 Pro keeps a person's face across new scenes. When in doubt, run the same prompt on two models and keep the better image.
Yes. Add photos as references and describe what to make with them. GPT Image 2.5 takes up to 16 images, Nano Banana up to 14 and Seedream 5.0 up to 10. To change one part of an image, open it in the image editor and type the change.
Yes. Put the exact words in quotes, like "LISBOA". GPT Image 2.5 Sunburst and Ideogram 4.0 are built for text. Check the spelling of every word before you publish.
Yes. Every paid plan includes commercial use, and images download without a watermark.
GPT Image 2.5, Nano Banana 2.1 and Pro, FLUX.3 and Kling 3.0 Omni go up to 4K. Most models offer shapes from 16:9 to 9:16, several go to 21:9, and Grok Imagine 2.0 goes from 2:1 to 1:2. Most make up to four images per run.
Sceneform blocks nudity, sexual content, undressing a person in a photo, and any sexual or revealing image of someone who looks under 18. A model can also decline a prompt on its own.
Type a sentence.Get the picture.
Generate an image
Photos, posters, illustration and edits.