For Developers/Models/GPT-Image-2.6 Flare
Closed SourceOpenAIReleased 2026-11

GPT-Image-2.6 Flare

OpenAI's fast GPT Image tier for quick iteration, now with GIF and SVG output

Context window

Not published

Input / 1M tokens

$8.00

Output / 1M tokens

$30.00

Provider

OpenAI

Per 1M tokens: text input $5.00 ($1.25 cached), image input $8.00 ($2.00 cached), image output $30.00, the same rates as GPT-Image-2.5. The listed $8/$30 is image input / image output. The Batch API is not supported, and there is one snapshot (gpt-image-2.6-flare). · Data verified 2026-11-08

GPT-Image-2.6 Flare (November 4, 2026) is the fast tier of OpenAI's Images 2.6 release, built for interactive generation and editing. It generates and edits images from text and image inputs, and the 2.6 models add a complexity setting, SVG output and animated GIF output. It costs the same as GPT-Image-2.5 Flare: $8 per 1M image input tokens and $30 per 1M image output tokens. Images 2.6 also runs in ChatGPT and Codex.

How to access it

Available in the OpenAI API (model ID gpt-image-2.6-flare) through the image generations and image edits endpoints; not through Chat Completions, Responses or Batch. Images 2.6 also powers image generation in ChatGPT and Codex.

Strengths

  • ✓Same per-token price as GPT-Image-2.5 Flare ($8 image input / $30 image output per 1M)
  • ✓Animated GIF and SVG output
  • ✓Markup-guided edits on the image edits endpoint
  • ✓Built for fast, interactive generation and editing

Best for developers who...

Fast, iterative image generation and editingApps where image latency mattersAnimated GIFs and SVG graphics

When to choose it (and when not to)

Reach for GPT-Image-2.6 Flare when...

  • →You want faster image generation and editing for interactive workflows
  • →You need animated GIF or SVG output from the API
  • →You use GPT-Image-2.5 Flare and want the newer model at the same price

Look elsewhere if...

  • ✕You need the most detailed compositions and retouching (use GPT-Image-2.6 Sunburst)
  • ✕You want Batch API discounts (not supported)

How to use it

  • ›Use markup-guided edits to mark the exact region you want changed
  • ›Draft with Flare and switch to Sunburst for detailed final images; both have the same per-token prices

Quickstart

Python
import base64
from openai import OpenAI

client = OpenAI()

result = client.images.generate(
    model="gpt-image-2.6-flare",
    prompt="A flat vector illustration of a lighthouse at dusk",
    size="1024x1024",
)
with open("lighthouse.png", "wb") as f:
    f.write(base64.b64decode(result.data[0].b64_json))

Image generations endpoint (v1/images/generations). Edits use v1/images/edits; GIF and SVG output are covered in OpenAI's image generation guide.

API model id: gpt-image-2.6-flare

Tools powered by GPT-Image-2.6 Flare

Compare GPT-Image-2.6 Flare

Compare GPT-Image-2.6 Flare with any other model

Build a comparison →
All model comparisons →

Learn the concepts