AI Image Model · OpenAI

GPT Image 2

GPT Image 2 is OpenAI's April 2026 image model, supporting both generation and editing through the API. It was succeeded by GPT Image 2.5 (September 8, 2026) but still ranked #3 on the LMArena text-to-image leaderboard on September 21, 2026, and remains a sensible default for blog covers, product scenes, ads, and diagrams.

At a glance (verified 2026-09-23)

Vendor
OpenAI
Type
AI Image Model
Pricing
OpenAI: $0.006 / $0.053 / $0.211 per ~1MP image (low / medium / high); Teamday static estimate $0.06 (medium)

What GPT Image 2 does for your business

Best for

General production imagesImage editingText-heavy graphicsMulti-element composition

How to use GPT Image 2 in Teamday

Available in Teamday for both generation and editing (up to 1536x1024) via media generation with an OpenAI key; Iris uses it as part of a mission rather than a one-off prompt.

Frequently asked questions

Can GPT Image 2 edit existing images?

Yes — it supports image generation and image editing through the OpenAI API, and Teamday's route supports both, including reference images for edits.

How much does GPT Image 2 cost per image?

OpenAI lists about $0.006 (low), $0.053 (medium), and $0.211 (high) per ~1-megapixel image; FAL charges the same, and Replicate lists $0.012 / $0.047 / $0.128. Teamday's static cost estimate is $0.06 per image at medium quality.

Is GPT Image 2 good at rendering text?

It is one of the leading candidates for text-heavy assets (alongside Gemini 3.1 Flash Image, Recraft V4.1, and Ideogram 4), but every public graphic still needs proofreading.

Which agents use GPT Image 2 in Teamday?

Iris, Teamday's visual designer, runs it from a brief — generating, editing, and saving approved assets with prompt and model provenance.

Related

GPT Image 2 is one of the models Teamday's AI employees already run — no API keys, no infrastructure.

Hire your first AI employee →