01
Generate from text
Build product frames, characters, environments and visual concepts from a written brief.
OpenAI GPT Image 2 in Saga
GPT Image 2, ready to generate and edit.
Use the current fast 1K route for new images or source-aware edits, with up to six references and a live dollar estimate before you render.
Saga concept visual for this workflow, not a GPT Image 2 benchmark
Saga availabilityLive now
Current Saga outputFast 1K
Reference imagesUp to 6
BillingPay per image
Live price and prompt
Check the meter, then direct the image.
Saga reads the current price from the same public endpoint used by the Studio. The prompt below is free to edit and copy. It is written for a clean product frame, but the structure works for people, environments and story stills too.
Current Saga estimate
Loading live quote...
GPT Image 2, fast 1K, per image
Preflight estimate. Completed image usage settles on your account ledger.
Free production prompt
What it does
Useful where a prompt alone is not enough.
OpenAI describes GPT Image 2 as its state-of-the-art model for image generation and editing. Saga exposes a focused fast route for rapid production work today.
02
Edit with sources
Send an existing image with a precise change request instead of restarting the composition.
06
Use references
Saga's current route accepts up to six reference images for identity, object and style direction.
05
Choose the frame
Work in 1:1, 4:3, 3:4, 16:9 or 9:16 without changing the model.
Saga route vs direct API
The limits are stated before the promise.
OpenAI's model documentation covers the full direct API. Saga's current route is intentionally narrower. This table separates the two so a visitor does not mistake a provider-wide capability for a setting available here.
| Capability | Saga now | OpenAI documentation |
|---|---|---|
| Generation | Available | Supported |
| Image editing | Available | Supported |
| Reference fidelity | Up to 6 inputs | High-fidelity inputs are documented for GPT Image 2 |
| Output size | Fast 1K route | Flexible sizes are documented for the direct API |
| Transparent background | Not offered | Not supported for GPT Image 2 |
Official references
Provider facts, linked directly.
Saga availability and limits come from the live Saga route. General model capabilities come from OpenAI's own documentation. The two are not blended together.
GPT Image 2 model page
Official model description, generation and editing capabilities, input fidelity and model snapshot.
Image generation guide
Official controls, supported workflow, output options and current limitations.
OpenAI API pricing
Official direct API pricing. Saga's live estimate above is separate and includes Saga's service margin.
Is GPT Image 2 available in Saga Studio?
Yes. Saga currently exposes GPT Image 2 for fast image generation and editing through its 1K route.
Does Saga offer 4K GPT Image 2 output?
Not on the current Saga route. Saga currently exposes fast 1K output. OpenAI documents flexible sizes for its direct API, but that is separate from the settings available in Saga.
Can GPT Image 2 edit reference images in Saga?
Yes. Saga's current route accepts up to six reference images for source-aware generation and editing.
Is the price exact?
The page fetches Saga's current preflight estimate. Image usage is settled from the completed output and recorded on your account ledger.
Does GPT Image 2 support transparent backgrounds?
No. OpenAI's current GPT Image 2 documentation says transparent backgrounds are not supported.
Generation and editing are live
Start in fast 1K. Keep the useful frame.
No subscription and no credit conversion. Saga shows the current dollar estimate before the render starts.
Use GPT Image 2