What the OpenAI roster covers
5
Video
2 spec pages
4
Image
1 spec page
The roster covers five capabilities. The deepest are edit image (3 models), image to image (3 models) and image to video (3 models).
| Capability | OpenAI models | Lowest headline price |
|---|---|---|
| Edit image | 3 | GPT Image 2 Image to Image · 2 cr |
| Image to image | 3 | GPT Image 2 Image to Image · 2 cr |
| Image to video | 3 | Sora 2 Image to Video · 10 cr |
| Text to image | 2 | GPT Image 2 Text to Image · 2 cr |
| Text to video | 2 | Sora 2 Text to Video · 10 cr |
OpenAI product lines
Three lines. Each one groups the models that share an underlying generation, so picking a line first and a tier second is usually faster than reading 9 rows.
Sora 2
5 modelsvideo · 10–50 credits · newest September 30, 2025
GPT Image 2
2 modelsimage · 2 credits · newest April 21, 2026
GPT Image 1.5
2 modelsimage · 2 credits · newest December 16, 2025
No spec page in this line — these models ship in the app but their catalog records carry no published specs.
All 9 OpenAI models
Every active OpenAI model in the Versely catalog, grouped by what it produces. Headline prices across the roster run from 2 to 50 credits. 5 of the 9 charge by the second of output, so those headline figures are rates rather than job totals — each model's own page carries the full range.
Video (5)
| Model | Line | Credits | Max output | Details |
|---|---|---|---|---|
| Sora 2 Image to Video | Sora 2 | 10/sec | 1080p | variant of Sora 2 Text to Video |
| Sora 2 Image to Video Pro | Sora 2 | 50 | 1080p | variant of Sora 2 Text to Video |
| Sora 2 Pro Storyboard | Sora 2 | 30/sec | — | Spec page |
| Sora 2 Text to Video | Sora 2 | 10/sec | 720p | Spec page |
| Sora 2 Text to Video Pro | Sora 2 | 50 | 1080p | variant of Sora 2 Text to Video |
Image (4)
| Model | Line | Credits | Max output | Details |
|---|---|---|---|---|
| GPT Image 2 Text to Image | GPT Image 2 | 2 | 4K | Spec page |
| GPT Image 2 Image to Image | GPT Image 2 | 2 | 4K | variant of GPT Image 2 Text to Image |
| GPT Image 1.5 | GPT Image 1.5 | 2 | 4K | In app only |
| GPT Image 1 Edit | GPT Image 1.5 | 2 | 4K | In app only |
Credit figures are each model's headline price; /sec, /MP and /1k chars mark the ones billed by output rather than by generation. 2 rows marked "in app only" have no spec page — the catalog record carries no specs to publish. Tier and mode variants link to the page that carries their row.
How OpenAI models are billed
Four different billing shapes across the roster — which is why a straight price comparison between two OpenAI models can mislead.
- Charged per second of output3 models
- Charged a flat rate per generation2 models
- Charged a flat rate that changes with output resolution2 models
- Charged per second, at a rate that changes with output resolution2 models
2 of the 9 currently carry a discounted credit price.
All prices are Versely credits. Your plan's credit allowance is on the pricing page.
OpenAI models with a live rank
4 of the 9 carry a rank in Versely's live model rankings. An unranked model is one the rankings have not measured — not a worse one. The 2 best-placed are below; the rest are in the roster table.
| Model | Rank | Elo | Price |
|---|---|---|---|
| GPT Image 2 Text to Image | #1 | 1370 | 2 credits |
| GPT Image 2 Image to Image | #2 | 1258 | 2 credits |
About the OpenAI roster
OpenAI's roster here is two families: Sora 2 for video and GPT Image for stills and edits.
Other model rosters
OpenAI model FAQ
How many OpenAI models does Versely have?+
9 active OpenAI models — 5 video models and 4 image models — documented across 3 spec pages. Tier and input-mode variants share a page with the model they are a variant of.
What is the cheapest OpenAI model?+
GPT Image 2 Text to Image carries the lowest headline price on the OpenAI roster at 2 credits. The highest is 50 credits. Compare those figures carefully: 5 of the 9 OpenAI models are billed by the second, so a headline number is a rate, not a job total.
Does OpenAI do more than video?+
Yes. The roster is 5 video models and 4 image models, covering edit image, image to image, image to video, text to image and text to video.
What is the newest OpenAI model?+
GPT Image 2 Image to Image is the most recently released OpenAI model in the catalog, dated April 2026. The oldest still active is GPT Image 1 Edit.
Are OpenAI models ranked?+
4 of the 9 OpenAI models carry a live Versely rank. The rest are listed by product line and price instead — an unranked model is one the rankings have not measured, not a worse one.
Why do some OpenAI models have no spec page?+
2 of the 9 are listed here without a link — GPT Image 1.5 and GPT Image 1 Edit. Their catalog records carry no specs beyond a one-line label, so a page for them would say nothing the table row does not. They are still available in the Versely app.
Run every OpenAI model on one subscription
The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync in a single app.