Generate from text
Write a prompt with gpt-image-2-t2i. Output photoreal portraits, product stills, lifestyle scenes, and concept art with explicit aspect ratio, quality tier, and resolution.
Loading...
Loading...
GPT Image 2 on OpenVideos
Photoreal text-to-image and natural-language photo editing
Use GPT Image 2 inside OpenVideos Create — new prompts default to gpt-image-2-t2i for text-to-image. Credits show before you submit; gallery samples below open Create with the same model and prompt.
OpenVideos exposes gpt-image-2-t2i for generation and gpt-image-2-i2e for edit workflows through the Model Registry. The credit table below lists default estimates — quality (low / medium / high), resolution (1k / 2k / 4k), and num_images update the final cost in Create before submit.
GPT Image 2 supports nine aspect ratios on t2i (1:1 through 21:9) plus auto on edit. Upload a photo in the hero form to route into gpt-image-2-i2e, or attach up to ten reference images in Create when you need multi-image edits or compositing.
Gallery samples on this page filter to GPT Image models only. Open View prompt for the full text, or Try sample to launch Create with the gallery preset.
Pair GPT Image 2 with our AI Image Generator hub for style spokes — product, portrait, and wallpaper — or jump to the image editor hub when your workflow starts from an existing photo.
Capabilities
One model pair covers generation and editing — all through the same Create panel and My Works download flow.
Write a prompt with gpt-image-2-t2i. Output photoreal portraits, product stills, lifestyle scenes, and concept art with explicit aspect ratio, quality tier, and resolution.
Upload a photo and pick gpt-image-2-i2e to replace backgrounds, adjust lighting, restyle outfits, or remove objects while preserving composition when you ask for it.
Attach multiple reference images in the editor workflow — useful for compositing, brand palette swaps, or keeping product geometry while changing scene context.
Default medium quality is about 40 credits at 1k; low and high tiers scale down or up, and 2k / 4k multiply the base. Browse chips in Create show the exact total before you spend.
How it works
Same Create modal and My Works flow as every OpenVideos image job — no separate generator page.
Type a prompt in the box above, start from a gallery sample, or open Create blank and paste from our inspiration gallery.
Default is gpt-image-2-t2i. Upload a reference in the hero form for gpt-image-2-i2e, or switch quality and resolution in Browse before submit.
Create shows the exact credit cost for quality, resolution, and image count. New accounts receive complimentary credits to try before buying packs on Pricing.
Finished images land in My Works — preview, regenerate with tweaks, or download. Failed jobs do not consume credits.
Prompt tips
Name lens, depth of field, and light source — e.g. “85mm portrait, shallow DOF, window light from camera left” — instead of only “realistic photo.”
Use low quality for fast layout drafts; medium for social-ready output; high when you need fine texture on products or skin.
gpt-image-2-i2e reads cleaner when the prompt describes a single adjustment — background swap, relight, or outfit change — not three unrelated edits.
When packaging, logos, or faces must stay on-model, start from the hero upload and describe only what should change in the scene.
Credit estimates
Estimates use default schema settings in Create. Quality, resolution, and image count change the final cost before submit.
| Model | From (credits) | |
|---|---|---|
| GPT Image 2 | 20 | |
| GPT Image 2 | 20 |
Gallery samples
Using the uploaded reference photo as the main subject, generate a high-end fashion collage style poster. Vertical 3:4 composition, the subject serving as the…
Related hub
Explore related tools →Related workflows
FAQ
This page pre-fills gpt-image-2-t2i in Create. Upload a reference in the hero form for gpt-image-2-i2e, or switch variants in Browse before submitting.
Cost depends on quality (low ~20, medium ~40, high ~100 credits at 1k), resolution multipliers for 2k and 4k, and num_images. Exact totals appear in Create before you submit.
Yes. Upload in the hero form for gpt-image-2-i2e, or open Create, choose image editor, and attach up to ten image_urls with your edit prompt.
gpt-image-2-t2i supports 1:1, 3:2, 2:3, 3:4, 4:3, 9:16, 16:9, and 21:9. The edit model adds auto aspect ratio for uploaded photos.
New OpenVideos accounts receive complimentary credits to try image and video generation. Paid credit packs are on the Pricing page — no subscription required for GPT Image 2.
Completed jobs appear in My Works. You can preview, download, or reopen a result in Create to iterate with the same GPT Image 2 model.
Plans
Use credits across AI video, image, short clip, and cinematic scene workflows. Choose the pack that matches your production volume.
New accounts get 50 free credits (30 days) — enough for about 5 AI images.
$12.90
1,000 credits · 30 days
$29.90
2,500 credits · 3 months
$79.90
10,000 credits · 6 months
$199.00
30,000 credits · 1 year
Secure checkout · Validity 30 days to 1 year
Ready to generate
Pick a brand page above or jump straight into the studio — credits apply per job, estimates show before submit.