> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ruapi.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# GPT Image 2.6 (Flare and Sunburst): GIF, SVG, pricing and what you can use from Russia

> GPT Image 2.6 launched on November 4, 2026: Flare vs Sunburst, how complexity, GIF and SVG work in the API, OpenAI pricing, and how to use GPT Image 2.5 from Russia without a VPN today.

On November 4, 2026 OpenAI released **ChatGPT Images 2.6**, along with two new API models: **GPT-Image-2.6 Flare** and **GPT-Image-2.6 Sunburst**. The headline features are animated GIFs, vector SVG and a separate `complexity` setting. Based on OpenAI's official materials, here is what changed since version 2.5, what it costs, and what you can use from Russia right now.

**In short**

* **What's new.** Images 2.6 in ChatGPT and Codex on every plan, plus the `gpt-image-2.6-flare` and `gpt-image-2.6-sunburst` models in the API.
* **Flare vs Sunburst.** Flare is the fast model for everyday generation (OpenAI says that with `quality: "low"` an image can be ready in as little as about six seconds). Sunburst is for detailed work: complex compositions and consistent details across edits.
* **New in the API.** A `complexity` parameter, GIF output (`output_format: "gif"`) and SVG output (`output_format: "svg"`). GIF and SVG are generation-only, not available for edits.
* **OpenAI's price.** The same rates as GPT Image 2 and 2.5: \$5 per million text input tokens, \$8 for image input, \$30 for image output. The Batch API is not supported for 2.6.
* **On RuAPI.** The 2.6 models aren't available yet. `gpt-image-2.5-flare`, `gpt-image-2.5-sunburst` and `gpt-image-2` are — from Russia, without a VPN or a foreign card.

## What OpenAI announced

The release is covered in the [OpenAI blog](https://openai.com/index/introducing-images-2-6/), the [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes) and the [API changelog](https://developers.openai.com/api/docs/changelog), all dated November 4, 2026 — less than two months after GPT Image 2.5 (September 8).

**In ChatGPT and Codex** (web, desktop and mobile, all plans):

* animated GIFs from a prompt or a sketch, with adjustable playback speed; the feature is rolling out gradually;
* saved reference images (yourself, pets, characters) that you can reuse in later chats without uploading them again.

**In the API** — two models for developers. OpenAI says Images 2.6 generates images 50% faster than before. Markup-guided edits (previously ChatGPT-only) and SVG export also come to the API.

## Flare vs Sunburst

| | GPT-Image-2.6 Flare | GPT-Image-2.6 Sunburst |
| - | - | - |
| API ID | `gpt-image-2.6-flare` | `gpt-image-2.6-sunburst` |
| Use case, per OpenAI | fast iteration and everyday generation, better layouts for posters and product images | detailed compositions and precise editing |
| Speed | per OpenAI, from \~6 seconds per image with `quality: "low"` | slower than Flare; OpenAI ties its quality gains to `quality` set to `medium` or `high` |
| Endpoints | `/v1/images/generations`, `/v1/images/edits` | `/v1/images/generations`, `/v1/images/edits` |
| GIF, SVG, `complexity`, markup | yes | yes |

The split is the same as in 2.5: Flare by default, Sunburst when precision matters. OpenAI warns that complex prompts can take up to 2 minutes.

## What's new in the API

### The `complexity` parameter

Separate from `quality` (visual detail), `complexity` sets how much the model plans the scene before rendering. Values: `low`, `medium`, `high`; if omitted, the 2.6 models use `high`. There is no separate fee, but higher values add latency and, as OpenAI warns, can introduce unwanted changes. For most edits OpenAI recommends `low`; for scenes with many elements or constraints, `medium` or `high`.

The docs describe `complexity` in the GPT Image 2.6 section; it isn't listed for the 2.5 models.

### Animated GIFs

* Request: `output_format: "gif"` on `POST /v1/images/generations`.
* Frame count via `frame_count` (default 9); frame duration via `frame_duration_ms` (default 200 ms; the guide gives 20–2000 ms, rounded to 10 ms).
* The `frame_count` limit differs between docs: the API reference says 2 to 16, the image generation guide says up to 64. Check what the API accepts on your account.
* No streaming, no partial images (`partial_images`), no compression setting and no `prompt_rewriter`.
* The API reference says GIF output requires access to the GIF preview for the selected model, so it may not be enabled on every account at first.

Describe the motion and the loop directly in the prompt.

### Vector SVG

* Request: `output_format: "svg"` on `POST /v1/images/generations`, `n=1` only.
* The SVG is written by a separate "author" model configured in the `svg` object: `svg.model` and `svg.reasoning.effort` (from `none` to `max`). Per the API reference, this requires access to SVG model selection; the reference doesn't list the allowed models.
* With SVG output, `complexity` controls the intermediate raster reference image, not the SVG author's reasoning.
* No streaming, partial images, compression or `prompt_rewriter`; the edit endpoint neither returns SVG nor accepts it as input.

> [!NOTE]
>
> OpenAI's docs don't spell out how many tokens the SVG author model uses — check the cost on your own tasks.

### Markup-guided edits

On `POST /v1/images/edits` you can pass an annotation layer on top of the image (`markup`), and the model changes the marked areas. Per the docs, this works on both the 2.5 and 2.6 models. Don't combine `markup` with a `mask` or `prompt_rewriter`.

## GPT Image 2.5 vs 2.6

| | GPT Image 2.5 (Flare / Sunburst) | GPT Image 2.6 (Flare / Sunburst) |
| - | - | - |
| Released | September 8, 2026 | November 4, 2026 |
| `quality` | `low`, `medium`, `high`, `xhigh`, `max` | `low`, `medium`, `high`, `xhigh`, `max` |
| `complexity` | not listed in the docs | `low`, `medium`, `high` (default `high`) |
| GIF and SVG | no | yes, generation only |
| Markup-guided edits | yes | yes |
| Where available | Image API and the Responses API image generation tool | Image API (`/v1/images/generations`, `/v1/images/edits`) |
| Batch API | yes, at a 50% discount | not supported |
| Price (per 1M tokens) | \$5 / \$8 / \$30 | \$5 / \$8 / \$30 |

## Pricing

OpenAI's official prices per 1M tokens ([OpenAI pricing page](https://developers.openai.com/api/docs/pricing), as of November 5, 2026), identical for `gpt-image-2`, both 2.5 models and both 2.6 models:

| | Input | Cached input | Output |
| - | - | - | - |
| Text | \$5 | \$1.25 | — |
| Images | \$8 | \$2 | \$30 |

In the Batch API (only `gpt-image-2` and the 2.5 models) rates are halved: \$2.50 / \$0.625 for text, \$4 / \$1 for image input, \$15 for image output. OpenAI notes that cached input rates for GPT Image 2 and 2.5 only apply to images generated with the Responses API.

The cost of a single image depends on its size and `quality`: the larger and more detailed the image, the more output tokens.

RuAPI prices per 1M tokens, as shown on the [pricing page](https://www.ruapi.ai/pricing?utm_source=docs\&utm_campaign=en-gpt-image-2-6) (as of November 5, 2026; your account may have a lower price):

| Model | Input (text) | Output |
| - | - | - |
| `gpt-image-2.5-flare` | \$6 | \$36 |
| `gpt-image-2.5-sunburst` | \$6 | \$36 |
| `gpt-image-2` | \$6 | \$36 |

Input images (for example, when you edit a picture) are billed separately at a higher rate than text input — the same way OpenAI does it.

## How to use it from Russia: GPT Image 2.5 on RuAPI

The 2.6 models aren't on RuAPI yet — we'll post in our [Telegram channel](https://t.me/ruapi_ai) when we add them. Both 2.5 models already work, from Russia without a VPN. Requests use the standard OpenAI format at `https://www.ruapi.ai/v1/images/generations`.

```bash curl theme={null}
curl -sS https://www.ruapi.ai/v1/images/generations \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -o resp.json \
  -d '{
    "model": "gpt-image-2.5-flare",
    "prompt": "A minimalist coffee shop logo: a cup with steam shaped like a cloud",
    "size": "1024x1024",
    "quality": "medium",
    "n": 1
  }'

tr -d '\\' < resp.json | grep -o '"b64_json": *"[^"]*"' | head -n 1 | cut -d'"' -f4 | base64 --decode > image.png
[ -s image.png ] && echo "Saved: image.png" || cat resp.json
```

```python Python (openai SDK) theme={null}
import base64

from openai import OpenAI

client = OpenAI(api_key="sk-YOUR_KEY", base_url="https://www.ruapi.ai/v1")

result = client.images.generate(
    model="gpt-image-2.5-flare",  # or gpt-image-2.5-sunburst
    prompt="A minimalist coffee shop logo: a cup with steam shaped like a cloud",
    size="1024x1024",
    quality="medium",
    n=1,
)

with open("image.png", "wb") as f:
    f.write(base64.b64decode(result.data[0].b64_json))
```

* `model` — `gpt-image-2.5-flare` for fast generation, `gpt-image-2.5-sunburst` when you need precision.
* `quality` — `low`, `medium` or `high`; it directly affects the number of output tokens, i.e. the price.
* `size` — e.g. `1024x1024`, `1536x1024`, `1024x1536`.

To edit existing images (with an `image` file and, optionally, a `mask`), use `https://www.ruapi.ai/v1/images/edits`. A sample response, parameters and error troubleshooting are on the [Image generation](https://docs.ruapi.ai/en/image-generation.md) page.

> [!WARNING]
>
> GIF and SVG are 2.6 features, so you can't get them through RuAPI yet. RuAPI currently does not forward the `complexity`, `frame_count`, `frame_duration_ms` and `svg` parameters or a `markup` file — don't include them in your requests.

Payment is in USDT (from 5 USDT; TRC20, BEP20, TON, Polygon networks; USDC is accepted too) or with a card from a foreign bank. No USDT? [Buy it for rubles via SBP](https://docs.ruapi.ai/en/buy-usdt.md).

## GPT Image vs Nano Banana

RuAPI also offers Google's Gemini Image models (Nano Banana). The main difference is how you call them:

* **GPT Image** is called via `/v1/images/generations` and `/v1/images/edits`, returns the image in `b64_json`, and is billed per token.
* **Nano Banana** is called via `/v1/chat/completions`, like a regular chat, and returns the image inside the response text; some models have a fixed per-image price, others are billed per token.

Both families and all their models, with examples, are covered on the [Image generation: Gemini Image and GPT Image](https://docs.ruapi.ai/en/image-generation.md) page.

> [card] [Sign up and try GPT Image](https://www.ruapi.ai/register?utm_source=docs\&utm_campaign=en-gpt-image-2-6)
>
> Sign up with email or with Google / Yandex. No VPN or foreign card needed.

## FAQ

**When will GPT Image 2.6 be available on RuAPI?**
There's no date yet. We'll announce `gpt-image-2.6-flare` and `gpt-image-2.6-sunburst` in our [Telegram channel](https://t.me/ruapi_ai) and in the [changelog](https://docs.ruapi.ai/en/changelog.md) when we add them.

**What's the difference between Flare and Sunburst?**
Flare is faster and fits most tasks. Sunburst takes longer but handles complex scenes more precisely and keeps details consistent across edits. The per-token price is the same.

**Can I get GIF or SVG from GPT Image 2.5?**
No. According to OpenAI's docs, GIF and SVG arrived with the 2.6 models only, and only for generation, not edits.

**Is GPT Image 2.6 more expensive?**
OpenAI's per-token rates haven't changed: \$5 / \$8 / \$30 per million tokens, the same as GPT Image 2 and 2.5. However, the 2.6 models don't get the Batch API discount.

**Do I need a VPN to use GPT Image from Russia?**
Not through RuAPI: requests go to `https://www.ruapi.ai/v1`, and you don't need an OpenAI key or a foreign card. SDK setup uses the same `base_url` as the GPT text models — see [GPT API](https://docs.ruapi.ai/en/gpt-api.md).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.