OpenAI
GPT Image 2.5 Flare
Fast text-to-image and editing with legible in-image text
imagelive
- Model id
- gpt-image-2.5-flare
- Typical run
- ~17s
- Inputs
- 9 parameters
- Contract
- async
Call it
bash
curl -X POST https://api.ezflows.io/v1/models/gpt-image-2.5-flare \
-H "Authorization: Bearer $EZFLOWS_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt": "a wide shot of a desert at dusk"}'
# { "request_id": "req_8f2a41c9", "poll_url": "…", "status": "queued" }Parameters
| Name | Type | Default | Description |
|---|---|---|---|
| prompt required | string | A photoreal product photograph of a froste… | Text describing the image. Supports in-image typography across scripts — lead with subject, style, lighting. |
| image_urls | string[] | — | Optional reference images to edit or draw context from. Providing any image switches to edit mode. |
| size | enum | 1024x1536 | Output resolution (WIDTHxHEIGHT). Each edge must be a multiple of 16, aspect ratio between 1:3 and 3:1, longest edge at most 3840, and total pixels between 655,360 and 8,294,400. "Auto" lets the model pick. |
| mask_image_url | string | — | Optional mask for surgical inpainting, used only alongside a reference image. Must be a PNG with an alpha channel, matching the reference image's dimensions: the transparent areas are regenerated, everything opaque stays pixel-perfect. A flat black-and-white mask with no alpha channel is rejected. |
| quality | enum | high | Rendering fidelity. "High" keeps typography crisp; use lower values only for previews. |
| background | enum | opaque | "Opaque" renders a filled background; "Transparent" returns a PNG or WEBP with an alpha channel; "Auto" lets the model decide. |
| moderation | enum | auto | Content filter strictness. "Auto" is the safe default. |
| output_format | enum | png | "PNG" for crisp text, "WEBP" for smaller size, "JPEG" for broad compatibility. |
| output_compression | integer | 100 | Compression level 0-100. 100 preserves text crispness; lower values reduce file size. |
What a run costs
$0.037average · per run
$0.00for a failed run
~17smeasured mean, end to end
How it stacks up
| Model | Price | Unit | Typical run |
|---|---|---|---|
| GPT Image 2.5 Flare | avg $0.037 | run | ~17s |
| Nano Banana 2 Lite | from $0.038 | image | ~12s |
| Seedream 4.0 (4k) | avg $0.035 | run | ~25s |
| Seedream 5.0 Lite: Image-to-Image | avg $0.035 | run | ~47s |
| Seedream 5.0 Lite: Text-to-Image | avg $0.035 | run | ~46s |