GPT Image 2.6 Sunburst vs P-Image-OSS
Newest: GPT Image 2.6 Sunburst · Self-hosting: P-Image-OSS · Most API hosts: GPT Image 2.6 Sunburst. Not enough arena votes yet for a quality score, so this is a spec comparison.
OpenAI
GPT Image 2.6 Sunburst
- Arena1,481
- From—
- Max size—
Pruna
P-Image-OSS
- Arena—
- From—
- Max size—
Add a model
Make it a three-way comparison.
Where they differ
Quality is the arena score from blind, side-by-side human votes. Price, length and resolution come from each lab’s own listings.
- NewestGPT Image 2.6 SunburstGPT Image 2.6 Sunburst Nov 4, 2026 · P-Image-OSS Nov 3, 2026
- Widest inputsSame inputsGPT Image 2.6 Sunburst: Text prompts, Reference images · P-Image-OSS: Text prompts, Reference images
- Image editingBoth edit imagesTakes a reference image as input: GPT Image 2.6 Sunburst yes · P-Image-OSS yes
- Self-hostingP-Image-OSSDownloadable weights
- Most API hostsGPT Image 2.6 SunburstGPT Image 2.6 Sunburst: 1 · P-Image-OSS: 0
Every spec in one table
Dashes mean the data is not published.
| Specification | ||
|---|---|---|
| Task | Not published | Image generation and editing |
| Published capabilities | Not published | Adapts the MiniMax H3 backbone to still-image generation and multi-reference edits. |
| Access | Not published | Public weights, ComfyUI workflow and Hugging Face demo |
| Operating limits | Not published | 2–16 megapixels and up to nine input images. The smallest published build is 20 GB. |
| Sources | Not published | Publisher documentation · Model weights |
| Quality (blind human votes) | ||
| Documented capabilities | Not published | Adapts the MiniMax H3 backbone to still-image generation and multi-reference edits. |
| Model parameters | Not published | Not published |
| Text to image arenaElo-style score · higher is better | 1,412 ±9#2 of 87 · 7,168 votes | Not published |
| Image editing arenaElo-style score · higher is better | 1,481 ±7#3 of 64 · 16,881 votes | Not published |
| Price | ||
| Starting price | Not published | Not published |
| Top price | Not published | Not published |
| Extra charges | Not published | Not published |
| How it is billed | Not published | Not published |
| Output | ||
| Produces | Images | Images |
| Resolutions | Not published | Not published |
| Highest resolution | Not published | Not published |
| Largest image | Not published | Not published |
| Images per request | Not published | Not published |
| Text in the reply | Not published | Not published |
| Aspect ratios | Not published | Not published |
| Inputs | ||
| Text prompts | Yes | Yes |
| Reference images (editing) | Yes | Yes |
| Max reference images | Not published | Not published |
| Access | ||
| Weights | Proprietary | Open |
| API model ID | Not published | Not published |
| API hosts | 1 | Not published |
| Released | Nov 4, 2026 | Nov 3, 2026 |
| Documentation | Publisher documentation | Publisher documentation · Model weights |
| Specification sources | Not published | Not published |
| Specs checked | Not published | Not published |
Which should you choose?
Which is better, GPT Image 2.6 Sunburst or P-Image-OSS?
Neither model has enough arena votes for a reliable quality score yet, so this page compares specs only. Run the same three or four prompts through each and judge the results.
What is the difference between GPT Image 2.6 Sunburst and P-Image-OSS?
GPT Image 2.6 Sunburst is an image model from OpenAI (Nov 4, 2026), taking text prompts and reference images and producing images. P-Image-OSS is an image model from Pruna (Nov 3, 2026), taking text prompts and reference images and producing images.
Which is newer?
GPT Image 2.6 Sunburst Nov 4, 2026; P-Image-OSS Nov 3, 2026.
Which can edit an existing image?
Takes a reference image as input: GPT Image 2.6 Sunburst yes; P-Image-OSS yes.
Can I run any of these on my own hardware?
P-Image-OSS. Downloadable weights.