GPT Image 2.6 API

GPT Image 2.6 is OpenAI's November 2026 image model and the successor to GPT Image 2.5. It keeps everything the previous tier does - text-to-image, editing from up to 16 reference images, quality from low to max - and adds two things. Set output_format: gif with frame_count (2-16) and you get a short animated loop. The new complexity field (low, medium, high) controls how much detail goes into the frame. The slower, more capable tier for long edit chains is GPT Image 2.6 Sunburst.

I tested the GIF output on a real call: a paper crane flapping its wings, 8 frames at 140 ms, 512x512. The loop is smooth and the file is about 300 KB. The same prompt at 1024x1024 with 6 frames was rejected, because all frames are drawn on one sheet and the sheet edge cannot exceed 3840 px. For GIFs, stay at 512x512.

Best for

• Animated stickers, icons and loops for email, chat and web (GIF)
• Covers, thumbnails and ad creatives at 1536x1024
• Posters with a headline - Polish diacritics came out right in my test
• Swapping the background or product in a photo while keeping the rest

Prompting tips

• For GIFs, describe one simple looping motion ("a paper crane flapping its wings in a seamless loop") on a flat background.
• GIF settings that worked: size: 512x512, 4-8 frames, frame_duration_ms around 120-160.
• complexity: low gives sparse, poster-like frames; high gives rich detail.
• Put any exact text in quotes and check it on the result.

FAQ

How much does the GPT Image 2.6 API cost?

Jantaris charges $0.0575 per image, taken from a prepaid balance, with no subscription. A GIF costs the same as one image. Failed generations are not charged.

How do I make an animated GIF?

Set output_format to gif, size to 512x512 and frame_count between 2 and 16. See the generate docs or try it in Studio.

Does GPT Image 2.6 output SVG?

Not on Jantaris yet. In my test OpenAI rejected the SVG format for this model, so I will add it only once it works on a real call.

What changed from GPT Image 2.5?

Same inputs plus GIF and complexity. On Jantaris it is also cheaper per image than GPT Image 2.5.

How to call GPT Image 2.6 via API

curl https://jantaris.com/v1/generate?wait=60 \
  -H "Authorization: Bearer $JANTARIS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "gpt-image-2.6",
  "input": {
    "prompt": "Product photo of a matte black ceramic coffee cup on a travertine table, soft window light, 50mm",
    "aspect_ratio": "auto"
  }
}'

import os, time, requests

API = "https://jantaris.com/v1"
H = {"Authorization": f"Bearer {os.environ['JANTARIS_API_KEY']}"}

task = requests.post(f"{API}/generate", headers=H, json={
    "model": "gpt-image-2.6",
    "input": {
        "prompt": "Product photo of a matte black ceramic coffee cup on a travertine table, soft window light, 50mm",
        "aspect_ratio": "auto"
    }
}).json()
while task["status"] in ("queued", "running"):
    time.sleep(3)
    task = requests.get(f"{API}/tasks/{task['id']}", headers=H).json()

print(task["status"], task["cost"], [o["url"] for o in task["output"]])

const API = "https://jantaris.com/v1";
const headers = { Authorization: `Bearer ${process.env.JANTARIS_API_KEY}`, "Content-Type": "application/json" };

let task = await (await fetch(`${API}/generate`, {
  method: "POST",
  headers,
  body: JSON.stringify({
    "model": "gpt-image-2.6",
    "input": {
      "prompt": "Product photo of a matte black ceramic coffee cup on a travertine table, soft window light, 50mm",
      "aspect_ratio": "auto"
    }
  }),
})).json();

while (task.status === "queued" || task.status === "running") {
  await new Promise((r) => setTimeout(r, 3000));
  task = await (await fetch(`${API}/tasks/${task.id}`, { headers })).json();
}
console.log(task.status, task.cost, task.output.map((o) => o.url));

Parameters

The same schema is available from GET /v1/models/gpt-image-2.6.