Three models
GPT Image 2.6 and GPT Image 2, plus Seedream 5 Pro.
AI Image Generator
Frontier image models in one panel — posters, product shots, edits, and concept frames up to 4K.
AIモデル 30 クレジット
プロンプト0/2000
著作権で保護されたキャラクター・ロゴ・音楽や、敏感・露骨な内容は避けてください。ブロックされる場合があります。
解像度
アスペクト比
長さ · 8s
バッチ生成1
推定コスト：50クレジット
作品がここに表示されます
モデルを選び、ショットを描写して「生成」を押してください。
AI Image Generatorの魅力
クリエイターがAI Image Generatorを選ぶ理由
Frontier image engines, one prompt box, and a clear credit estimate before every render.
Up to 4K
Export resolution suited to web and print.
Edit as well as create
Upload references to change existing images.
Clear costs
See the estimate before you generate.
AI image generator guide
How to use the AI image generator
This AI image generator brings frontier image models into one panel: GPT Image 2.6, GPT Image 2, and Seedream 5 Pro. You can start from a prompt or from reference images and export up to 4K. This guide explains how to pick a model and a workflow.
What you can create
Typical jobs include posters, product shots, social graphics, concept frames, and edits to existing photos. Every model supports text to image and image to image, and exports at up to 4K. The panel shows the credit estimate before each run.
Choosing an image model
Use GPT Image 2 for fast drafts and everyday volume. Use Seedream 5 Pro for multi-reference composites and consistent subjects across up to ten inputs. Use GPT Image 2.6 for text-heavy designs such as posters and packaging and for native 2K output.
Speed and volume
GPT Image 2.
Multi-reference fidelity
Seedream 5 Pro.
Typography
GPT Image 2.6.
A workflow that saves credits
Explore at 1K, choose a direction, then render the final at 2K or 4K. If a generation fails on our side, credits are refunded automatically. New accounts get free credits, and paid plans include commercial usage rights.
AI Image Generatorの使い方
AI Image Generatorを3ステップで使う
Describe or upload
Write a prompt or add reference images.
Select the model
Pick the engine that fits speed, fidelity, or text.
Generate and download
Compare variations and export the final.
AI Image Generatorのユーザー
画像制作にAI Image Generatorを使う人々
Illustrators
Concept art and reference frames.
Marketing teams
Campaign visuals for many placements.
Store owners
Product images and lifestyle scenes.
Everyday users
Invitations, avatars, and personal projects.
その他のモデル
AI Image Generatorと他の全モデルを1つのスタジオで
Switch engines without switching tools — Gemini Omni 1.2 Flash plus strong public image and video models, side by side in the Gemini Omni 1.2 Flash studio.
Text-to-videoReference-guided shotsUp to 7 image references720p–4K output16:9 & 9:164–10s clipsFast Flash draftsBatch runsGPT Image key visuals
Useful answers before you create
AI Image Generatorに関するよくある質問
Understand the model, prepare better references, and choose a practical approach to your next clip.
01Which image models are included?
GPT Image 2.6 — our flagship for native 2K output and faithful editing — plus GPT Image 2 and Seedream 5 Pro.
02What resolution can I export?
Up to 4K on the image models —1K drafts are cheapest.
03Can I edit existing images?
Yes — Image to Image mode accepts reference uploads for edits, composites, and variations.
04Which model should I pick?
GPT Image 2.6 for text-heavy designs, product edits, and native 2K; GPT Image 2 for fast everyday drafts; Seedream 5 Pro for multi-reference composites.
Make the next shot visible
Start your next clip on Gemini Omni 1.2 Flash
Bring a character still, a product photo, or a clear shot brief. Draft a take, refine the direction, and save the clip that fits your edit.