GPT-Image-2.5 Flare Specifications
The two variants share much of the same API surface, but their optimization targets are different.
| Specification | Flare |
|---|---|
| Model ID | gpt-image-2.5-flare |
| Snapshot | gpt-image-2.5-flare-2026-09-08 |
| Primary positioning | Fast, high-quality everyday generation |
| Inputs | Text, images |
| Output | Images |
| Image generation | Yes |
| Image editing | Yes |
| Recommended default role | Most applications |
| Quality modes | low, medium, high, xhigh, max, auto |
| Recommended square size | 1024×1024 |
| Recommended landscape size | 1536×1024 |
| Recommended portrait size | 1024×1536 |
| Custom dimensions | Yes |
| Maximum edge | 3840 px |
| Maximum total pixels | 8,294,400 |
| Supported aspect-ratio range | 1:3 to 3:1 |
| Transparent background | PNG / WebP |
| Official text input price | $5 / 1M tokens |
| Official image input price | $8 / 1M tokens |
| Official image output price | $30 / 1M tokens |
What Is GPT-Image-2.5 Flare?
GPT-Image-2.5 Flare is OpenAI's fast, general-purpose image-generation and editing model in the GPT Image 2.5 family. It is designed for applications where image quality and generation speed both matter, rather than maximizing editing precision at the expense of latency.
OpenAI positions Flare as the default choice for most applications, including creator and social content, product experiences, visual search, rapid image prototyping, and high-volume image generation. Compared with GPT Image 2, OpenAI reports higher image quality with up to 50% lower latency.
Main Features of GPT-Image 2.5 Flare
- Fast image generation: Optimized for high-quality everyday generation with lower latency than GPT Image 2.
- Text-to-image generation: Generates images directly from natural-language prompts.
- Image-to-image editing: Accepts image inputs for editing and transformation workflows.
- Improved visual fidelity: OpenAI reports more natural lighting, richer textures, and sharper visual details in the GPT Images 2.5 generation.
- Better reference preservation: Improved at preserving subjects from reference images during transformations and edits.
- Complex layouts and transparency: The model supports image-generation workflows involving detailed layouts and transparent backgrounds.
- Flexible quality control: Supports six quality settings:
low,medium,high,xhigh,max, andauto.
Benchmark Performance of GPT-Image 2.5 Flare
The strongest currently verifiable performance statement is OpenAI's claim that Flare provides higher-quality images than GPT Image 2 with up to 50% lower latency.
| Model | Text-to-Image Elo | Single-Image Edit Elo |
|---|---|---|
| Sunburst | 1421 ± 13 | 1520 ± 9 |
| Flare | 1399 ± 13 | 1491 ± 9 |
| Image 2 (medium) | 1381 ± 4 | 1461 ± 3 |
| MAI Image 2.6 | 1331 ± 7 | 1439 ± 8 |
| Grok Imagine Image 2.0 (low) | 1315 ± 12 | 1434 |
As of September 9, 2026.
GPT-Image-2.5 Flare vs GPT Image 2 vs GPT-Image-2.5 Sunburst
| Model | Primary positioning | Best suited for |
|---|---|---|
| GPT-Image-2.5 Flare | Fast, high-quality everyday image generation | High-volume generation, social content, product experiences, rapid iteration |
| GPT Image 2 | Previous-generation general image model | Existing GPT Image 2 integrations |
| GPT-Image-2.5 Sunburst | Higher precision image generation and editing | Production creative, detailed editing, workflows where edit precision matters most |
The key distinction within the GPT Image 2.5 family is speed vs. precision: OpenAI positions Flare for fast, high-quality everyday generation, while Sunburst is intended for workflows that need tighter control over detailed creative work.
Limitations
GPT-Image-2.5 Flare should not be described as a general multimodal language model. Its documented modalities are text/image input and image output; audio and video are unsupported. It also does not provide function calling, structured outputs, fine-tuning, or streaming.
For benchmark claims, avoid unsupported numerical scores.
How to Access GPT-Image-2.5 Flare API with CometAPI
CometAPI currently lists gpt-image-2.5-flare with both /v1/images/generations and /v1/images/edits endpoints. Its model comparison page also confirms the September 2026 release and the API model ID.
Step 1: Get a CometAPI API Key
Create or sign in to your CometAPI account and obtain an API key from the developer console.
Step 2: Select the GPT-Image-2.5 Flare Model
Use:
gpt-image-2.5-flare
Then send your image-generation or image-editing request through the corresponding CometAPI image endpoint.
Step 3: Process the Generated Image
The API returns the generated image result. Your application can then store, display, or further process the image according to your workflow.