GPT Image 2.5 Sunburst Text to Image API

openai/gpt-image-2.5-sunburst/text-to-image
1K · 2K · 4K

GPT Image 2.5 Sunburst Text to Image turns complex prompts into visuals up to 4K, with fine material textures and realistic lighting. It accurately reproduces depth and fine detail while maintaining precise composition under long, detailed instructions.

Get API Key
Input
Required1219/32000
Optional
Optional
Optional
Optional
Optional
OutputReady

Estimated cost: high / 2K · 23.94 credits · $0.11970

Billed per image, based on the selected quality and resolution. 1 credit = $0.005.

Continue using

Examples

output-a1ad5d9c3d14.png

Create a photorealistic intimate backstage portrait after a theatre performance. One adult female ballet dancer in her early thirties sits on a worn wooden chair in the foreground, shoulders relaxed, looking down with a tired but relieved expression. She wears a muted lavender silk practice dress; satin pointe shoes rest beside her bare feet on the floor. A tall dressing-room mirror behind her reflects the same woman's back and a narrow warm spill of light from a distant stage doorway, with geometrically correct reflection and no duplicate faces. Cool dim ambient light in the room, a few warm bulbs above the mirror. Natural skin and subtle perspiration, fine silk folds and chair grain, stillness and emotional restraint. No costumes for sale, posters, writing or additional people. No brands, logos, products for sale, advertisements, promotional graphics, watermark, split screen or comparison panels.

output-bdbf0bbd5613.png

Create a cinematic photorealistic wide view of an underground public archive moments after power returns. A middle-aged male archivist in a plain burgundy cardigan pushes a narrow steel trolley stacked with closed document folders along the central aisle, moving away from camera but turning his head slightly toward the newly lit shelves. Tall dark metal shelves stand in two straight parallel rows with clear one-point perspective toward a distant stone doorway. Foreground shelves remain dim; three ceiling lamps farther down the aisle are lit, forming distinct pools of warm light through fine dust in the air. One simple sign above the far doorway reads exactly "ARCHIVE 07" in large legible uppercase letters; no other writing is visible. Rich paper edges, aged folder cloth, brushed trolley metal and rough subterranean stone. Coherent wheels on the floor, realistic hand positions, no floating folders, no crowds, no split layouts. No brands, logos, products for sale, advertisements, promotional graphics, watermark, split screen or comparison panels.

GPT Image 2.5 Sunburst Text to Image

GPT Image 2.5 Sunburst Text to Image is OpenAI’s flagship model for precise image generation, ranking first in Arena blind evaluations. Built for exceptional image quality and commercial finals, it natively supports up to 4K resolution, transparent backgrounds and multiple quality levels. Its strengths in complex prompt adherence, interactions between subjects, fine lighting, skin and fabric rendering meet the visual demands of advertising and publishing.

Why Choose GPT Image 2.5 Sunburst Text to Image?

  • Leading image quality in Arena evaluationsRanked first in text-to-image blind evaluations with an Elo score of 1421, leading in prompt adherence, visual aesthetics and realistic complex scenes.

  • Fine material detail and realistic lightingRender skin pores, fabric fibers, glass refraction and brushed metal with fine detail, realistic depth of field and natural, clear lighting.

  • Precise spatial logic and long-prompt adherencePrompts of up to 32,000 characters let you specify relative character positions, complex prop interactions and precise spatial perspective for the model to reproduce.

  • Native resolution up to 4K and fine quality controlNative 1K, 2K and 4K output adds premium xhigh and max levels alongside the familiar low through high options, supporting large-format commercial printing.

  • Transparent output and flexible aspect ratiosGenerate PNG or WebP assets with transparent alpha channels directly, with 10 common preset aspect ratios and custom size support.

Parameters

ParameterRequirementDescription
promptRequired

Non-blank text, up to 32,000 characters.

sizeOptional

Default 1:1. Ratios: auto, 1:1, 2:3, 3:2, 4:3, 3:4, 4:5, 5:4, 16:9, 9:16, 21:9. auto requires 1K. Custom WIDTHxHEIGHT requires 2K/4K, positive multiples of 16, maximum edge 3840, ratio at most 3:1 and 655360–8294400 pixels; a 3840 edge requires 4K.

Default1:1
resolutionOptional

1K, 2K or 4K. Default 1K.

Default1K2K4K
output_formatOptional

png, jpeg or webp. Default png.

Defaultpngjpegwebp
qualityOptional

low, medium, high, xhigh or max. Default high. Price depends on quality and resolution.

Defaulthighlowmediumxhighmax
backgroundOptional

auto, transparent or opaque. Default auto. transparent requires png or webp.

Defaultautotransparentopaque

How to Use

  1. Describe your visual scene in detailUse prompt to describe the subject, material textures, depth of field, light direction and artistic style. Mark any text slogans with double quotation marks.

  2. Set the resolution and aspect ratioChoose a preset such as 1:1, 16:9 or 21:9, or enter custom WIDTHxHEIGHT dimensions, then select 1K, 2K or 4K resolution for your deliverable.

  3. Choose the output format and qualitySelect high, xhigh or max according to your quality requirements. For transparent design assets, set background to transparent.

  4. Submit an asynchronous generation requestSubmit your parameters through the API to receive a unique task_id. The high-performance cluster starts detailed spatial reasoning and precise rendering.

  5. Retrieve and download the finished imagePoll the task status using task_id. Once the status is finished, read the returned high-resolution image URL and save your result.

Usage and Pricing

Each request generates one image. Pricing depends on quality and resolution, starting at 1.31 credits per image. The API default high / 1K setting costs 11.78 credits per image ($0.05890). 1 credit = $0.005. See the table below for all rates.

Usage and Pricing
QualityResolutionCost per image
low1K
1.31 credits
$0.00655
2K
2.65 credits
$0.01325
4K
2.47 credits
$0.01235
medium1K
2.95 credits
$0.01475
2K
5.99 credits
$0.02995
4K
5.81 credits
$0.02905
high1KAPI Default
11.78 credits
$0.05890
2K
23.94 credits
$0.11970
4K
22.39 credits
$0.11195
xhigh1K
20.93 credits
$0.10465
2K
42.55 credits
$0.21275
4K
39.78 credits
$0.19890
max1K
47.11 credits
$0.23555
2K
95.74 credits
$0.47870
4K
89.51 credits
$0.44755

Best Use Cases

  • Commercial advertising and campaign postersCombine detailed lighting and 4K resolution to produce final assets for large outdoor billboards and digital displays without additional retouching.

  • Film storyboards and concept environmentsUse strong long-prompt understanding to reproduce expansive scenes with complex architecture, dramatic lighting and interactions between characters.

  • Luxury and consumer electronics visualsReproduce matte metal, jewel refraction, leather textures and clear glass for premium product catalogs and ecommerce hero images.

  • Book covers, editorial illustration and digital artCreate accomplished compositions with complex artistic styles and rich detail for high-fidelity publishing and premium artwork.

Pro Tips

  • Describe materials and light precisely: phrases such as "brushed metal with a shallow macro depth of field" or "early morning window light from a 45-degree rear side angle" help bring out Sunburst’s rendering capabilities.
  • Use max for commercial finals: for large-format prints or demanding commercial output, combine 4K resolution with max quality for fine pixel-level detail and smooth gradients.
  • Use a transparency-compatible format: transparent backgrounds require png or webp. jpeg does not support an alpha channel and will fail validation.
  • Structure long prompts: with up to 32,000 characters available, organize descriptions by subject appearance, spatial layout, material details, then lighting and color so the model can parse complex constraints.

Notes

  • prompt is required and supports up to 32,000 characters, allowing the model to interpret complex composition constraints.
  • Pricing depends on five quality levels from low to max and three resolution options: 1K, 2K and 4K. Generation starts at 1.31 credits per image.
  • Sunburst focuses on precision and realistic materials. Inference for a single image takes slightly longer than the speed-focused Flare variant.
  • Submission returns a task_id. Retrieve results through standard asynchronous polling and ensure your polling logic handles network issues reliably.

Related Models

GPT Image 2.5 Sunburst Text to Image API FAQ

What is GPT Image 2.5 Sunburst Text to Image API?

GPT Image 2.5 Sunburst Text to Image is OpenAI’s flagship model for precise text-to-image generation. It produces commercial-quality images up to 4K from text prompts, ranks first in Arena blind evaluations, and offers fine material rendering, realistic lighting and multiple quality levels. Built on OpenAI’s advanced spatial and physical understanding, it closely follows complex, lengthy prompts while producing realistic depth of field and nuanced textures. You can call it programmatically through the API or try it directly in the Playground above.

What image quality and detail advantages does GPT Image 2.5 Sunburst offer over Flare?

Sunburst is built for demanding commercial deliverables and final artwork. It reaches OpenAI’s highest level of image generation in realistic material textures, complex perspective, fine light refraction and detailed typography, making it suitable for premium design and film concepts with exacting detail requirements.

When should I use max quality with GPT Image 2.5 Sunburst Text to Image?

max is Sunburst’s ultimate quality option. It uses more computation steps to refine fine pixel textures and complex lighting transitions, making it well suited to large advertising prints, luxury product catalogs and high-specification art publications.

How does GPT Image 2.5 Sunburst Text to Image handle prompts of up to 32,000 characters?

The model optimizes contextual attention for long natural language descriptions. It extracts character traits, prop details, lighting arrangements and style instructions from lengthy text, avoiding the forgotten elements or confused combinations common with traditional models.

Can GPT Image 2.5 Sunburst Text to Image generate transparent assets directly?

Yes. Set background to transparent and output_format to png or webp. The model identifies the foreground subject and generates an asset with a smooth, antialiased alpha channel, ready to use without background removal.

Does GPT Image 2.5 Sunburst Text to Image cost the same as Flare?

Yes. Sunburst and Flare share the same 5×3 pricing matrix, based on quality and resolution. low / 1K costs 1.31 credits ($0.00655) per image, and the default high / 1K costs 11.78 credits ($0.05890). There is no additional premium for Sunburst’s flagship precision.

What restrictions apply to custom sizes in GPT Image 2.5 Sunburst?

Pass a WIDTHxHEIGHT string through size. Both width and height must be positive multiples of 16, the longest edge must not exceed 3840 pixels, the aspect ratio must not exceed 3:1, and the total pixel count must be between 655360 and 8294400. Custom sizes require 2K or 4K resolution.