GPT Image 2.5 Sunburst API
GPT Image 2.5 Sunburst is OpenAI's quality-first base model for the API. It offers higher fidelity than GPT Image 2 for brand assets, polished product imagery, and multi-pass refinement. Roko exposes generation and reference-image editing at $0.02 per 1K image, $0.03 per 2K, and $0.04 per 4K.
- Public model ID
gpt-image-2.5-sunburst- Output tiers
1K / 2K / 4K (1K default)- Billing
Per output by resolution- Endpoint
POST /v1/images/tasks- Catalog snapshot
What is GPT Image 2.5 Sunburst?
GPT Image 2.5 Sunburst is one of the two API image models OpenAI shipped with the ChatGPT Images 2.5 launch. Officially it is the “base model, optimized for quality”: higher image quality than GPT Image 2, aimed at production-ready campaign creative and polished product imagery. Early customers praise that it “understands what not to change” — making meaningful edits while preserving character, composition, and visual identity.
OpenAI supports generation, editing, and transparent backgrounds upstream. Roko's current public form exposes generation and reference-image editing but no background or output_format control, so this page does not promise alpha-channel output.
Roko currently exposes 1K, 2K, and 4K tiers plus 14 size choices including auto at $0.02 / $0.03 / $0.04 per image. Calls run as async tasks — submit, then poll for the result.
Core capabilities
The official quality tier
A base model above GPT Image 2 in fidelity, giving demanding visual workflows finer control.
Precise editing, subject preservation
OpenAI and early customers highlight that it knows what not to change: edit what matters while keeping character and composition.
Upstream transparency
OpenAI supports transparent backgrounds; Roko does not currently expose the dedicated controls, so request validation remains authoritative.
Brand-ready output
Top-tier output plus $0.02 starting pricing fits brand assets, product imagery, and campaign finals.
GPT Image 2.5 Sunburst API pricing
Pricing is generated from the current Roko catalog and billed per output at the chosen resolution. Confirm the console estimate before submitting.
| Resolution | Price | Billing unit |
|---|---|---|
| 1k | $0.02 | per output |
| 2k | $0.03 | per output |
| 4k | $0.04 | per output |
Cost examples
Entry cost for brand review and precision composition.
Typical spend for a batch of finished brand or product assets; failed images are not billed.
Parameters and supported scope
These fields come from the current public generation_config. quality exposes auto only; capabilities outside the form follow request validation.
| Parameter | Values | Notes |
|---|---|---|
prompt | textarea | Describe the image to create or the changes to apply to references. · Required |
image_urls | url / base64 | Reference images may use URL or Base64. State what must stay and what may change. · Optional |
size | auto / 1:1 / 16:9 / 9:16 / 4:3 / 3:4 / 3:2 / 2:3 / 5:4 / 4:5 / 21:9 / 9:21 / 2:1 / 1:2 / 1:3 / 3:1 | Selects an aspect ratio, resolution tier, or explicit pixel dimensions as configured for the model. · Default: auto · Optional |
resolution | 1K / 2K / 4K | Selects an output resolution tier configured for this model; the value may directly affect price. · Default: 1K · Optional |
quality | auto | The quality tier follows the model config (auto / low / medium / standard / hd) and can directly affect price. · Default: auto · Optional |
Call GPT Image 2.5 Sunburst with Roko API
Run the async task flow: submit a task to get a task_id, then poll for the result. Never expose an API key in client-side code.
Submit generation
curl https://api.rokoapi.com/v1/images/tasks \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: YOUR_UNIQUE_REQUEST_ID" \
-d '{
"model": "gpt-image-2.5-sunburst",
"prompt": "Premium skincare bottle campaign image, restrained typography, accurate label text, soft studio light, preserve the brand mark",
"size": "16:9",
"resolution": "2K",
"quality": "auto"
}'Query recovery task
curl https://api.rokoapi.com/v1/tasks/$TASK_ID \
-H "Authorization: Bearer $API_KEY"POST /v1/images/tasks returns 202 with a task_id; once generation completes, query GET /v1/tasks/{task_id} for the result or failure reason. Idempotency-Key must be unique; replays with the same key return the original task without re-billing.
Best-fit scenarios
Campaign-ready brand art
Quality-first output for precise text, brand marks, and polished compositions.
Luxury watch campaign key visual, preserve the exact logo and dial details, dramatic controlled lighting.
Product-image refinement
Edit from a reference and name every preservation constraint for product and packaging iterations.
Replace only the background with warm travertine, keep the bottle, label, cap, and shadows unchanged.
Multi-pass local edits
Sunburst targets subject, layout, and visual-identity consistency across repeated changes.
Keep the character and layout unchanged; revise only the headline to “Autumn Collection”.
High-stakes marketing assets
Review at 1K from $0.02, then move approved finals to 2K or 4K when needed.
Polished launch poster for a premium speaker, clean grid, realistic materials, precise headline placement.
Important limitations
- It is OpenAI's model for exactly these brand-final, multi-pass workflows.
- quality currently exposes auto only; capabilities outside the form follow request validation.
- A submitted task is not a successful generation — trust the task query result.
- OpenAI applies provenance measures (C2PA metadata, invisible watermarking) to its image products; output rights depend on your inputs and applicable terms.
Flare, Sunburst, or GPT Image 2?
Pick Sunburst for top fidelity and consistency across repeated edits; pick Flare for everyday speed and volume; validated existing workflows can stay on GPT Image 2 ($0.016 per 1K is cheaper).
All models →GPT Image 2.5 Sunburst FAQ
What model ID should I use for GPT Image 2.5 Sunburst?
Use the public Roko API model ID gpt-image-2.5-sunburst exactly as shown in the model field.
How much does the GPT Image 2.5 Sunburst API cost?
The current catalog lists $0.02 per 1K output, $0.03 per 2K, and $0.04 per 4K. Prices can change; confirm the console estimate and final settlement.
Flare vs Sunburst — what's the difference?
Per OpenAI: Sunburst is the base model optimized for quality, above GPT Image 2; Flare is the small model optimized for speed with quality comparable to GPT Image 2. Both share generation, editing, and transparent backgrounds — the trade-off is speed versus fidelity.
When should I choose Sunburst?
Choose Sunburst when GPT Image 2 or Flare misses your fidelity, precise-editing, or multi-pass consistency bar. If Flare already meets it, keep the faster model and benchmark both on the same prompt set.
Does it support transparent backgrounds?
OpenAI supports it upstream, but Roko's current public form has no background or output_format field. This page therefore does not promise alpha-channel output; follow request validation and the actual response.
Which aspect ratios are supported?
The current form has 14 choices: auto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9, 9:21, 2:1, and 1:2. The default is auto; request validation is authoritative.
Which endpoint does it use?
Submit async tasks via POST /v1/images/tasks to get a task_id, then poll GET /v1/tasks/{task_id} for the result or failure reason.
What does migrating from GPT Image 2 involve?
Swap the model field to gpt-image-2.5-sunburst; Roko's current form is otherwise nearly identical. Compare the same prompts and references, and switch only when Sunburst provides a material fidelity or editing-stability gain.