릴리스노트

Fahsion AI - Changelog

총건수

Saved editorial cuts no longer end at "reference." Model Poses is the newest app in Abocado AI's Fashion AI — it moves a real model photo's pose onto your own photo. One photo of you, one model photo carrying the pose — two uploads, and the stance comes over. Your face, build, and the clothes you're wearing stay exactly as they are; only the pose transfers.

What's New

You make one choice: how much comes over.

  • Pose only — the stance alone. Everything down to the background stays your photo.

  • Pose + supports — the stance plus whatever it leans on (chair, wall, railing). The surroundings are still yours.

  • Pose + background — the model's whole scene, location and lighting included. What stays yours: face and outfit.

There are no prompts. The referenced model never appears in the result — the photo is read for stance only — and if your hand holds a cup or a bag, that hand keeps its item while the rest of the body follows the pose.

Who It Helps

Online sellers producing outfit cuts at volume, creators building a consistent feed, and anyone who wants to recreate a magazine pose as-is — no studio, no pose coaching, any editorial stance ready to use.

How It Differs From Templates

Picking from a fixed pose list means the list is the ceiling. Model Poses sources from any real photo, so there is none — editorials, lookbooks, street shots alike. Dress the look with Model TryOn, then vary the stance cut by cut with Model Poses, and a full editorial set comes together without a shoot.

Where to Find It

Create page, Fashion AI tab → Model Poses. Upload your photo + the pose model photo → pick a Pose Type (3 modes) → generate. The ratio auto-follows your photo's framing.

FAQ

Does the person in the model photo show up in the result?

No. The model photo is read for stance only — no face, clothes, or skin comes over. The person in the result is always you.

Do I need to describe the pose in words?

No. There is no text input for poses at all. Upload a photo carrying the pose you want and the app reads it directly — for professional poses, a photo is more precise than words.

Which of the three modes should I start with?

The smallest one that fits. Keep your background → pose only; a seated or leaning stance → pose + supports; the model's whole scene → pose + background.

The Deeper Story

From the principle of keeping identity and pose on two separate axes, to choosing between the three modes and hands-on tips — covered in two deep-dives.

Model Poses Intro — Three Modes and How They Work

Why only the pose moves while you stay you — the two-axis principle, with example cuts

Read More

Model Poses Guide — Choosing a Mode, Plus Hands-On Tips

Step-by-step usage, when to pick which mode, and the unlimited-editorial workflow

See the Guide

Explore Fashion AI

Browse all Fashion AI apps on Abocado AI

Explore

Start Creating

Two photos put that editorial pose on yours

Create

Until now, virtual try-on in Fashion AI started from composition — picking tops, bottoms, and outerwear from a look sheet and stacking them. Model TryOn starts from the opposite end. The finished look is the source. One full-outfit photo of a model you love, and that look lands on your photo wholesale. It's the newest app in Abocado AI's Fashion AI.

What's New

  • Model TryOn (base) — the model photo's entire outfit (clothes, shoes, accessories) onto your photo. Your face, pose, and background stay untouched; up to 4K.

  • Model TryOn Pose — outfit plus pose, together. Background is a card choice: My Background (your space stays; any support the pose needs comes along in shape only) / Model's Background (the editorial's location and light, recreated).

On both models, your face, height, and build stay yours, with garment size and length re-fitted to your body.

Who It Helps

Store operators with no time to produce look sheets, influencers who need editorial-mood profile and lookbook cuts, and anyone who wants to check "what if I dressed like that model?" right away — results first, no shoot or styling plan.

Model TryOn vs. TryOn Pro

  • Composing and layering items yourself → TryOn Pro

  • A finished look wholesale, up to pose and background → Model TryOn

Where to Find It

Create page, Fashion AI tab → Model TryOn. Upload your photo + the model photo → Model dropdown (TryOn/Pose) → (Pose) background mode card → generate. The ratio auto-matches your photo.

FAQ

I already use look sheets — should I switch?

No; they serve different jobs. Item-level planning and composition is still TryOn Pro's domain, while Model TryOn is fastest when you want a finished reference recreated wholesale. Moving between the two is the efficient workflow.

What makes a good model photo?

A full-body shot where the outfit reads head to toe. Accessories like sunglasses and bags transfer as part of the outfit, so skip references carrying items you don't want.

The Deeper Story

Identity preservation, two background modes (My / Model's Background), and support-object handling for poses — the design principles and step-by-step usage of Model TryOn are covered in two deep-dives.

Model TryOn Intro — How a Finished Look Transfers Wholesale

Model TryOn's design principles, identity preservation, and the story behind the two background modes

Read More

Model TryOn Guide — From Outfit Transfer to Background Recreation

Two photos → pick a model → background mode → generate, with step-by-step usage and photo tips

See the Guide

Explore Fashion AI

Browse all Fashion AI apps on Abocado AI

Explore

Wear It Now

Two photos put that model's look on yours

Try On

We've stabilized the 1.0 Pro output of Model-Gen, our virtual fashion model generation app. As the generation engine's content safety policy was tightened, some results were occasionally not generated — we've addressed this, so you can now receive model images more reliably while the person stays exactly the same.

What changed

Model-Gen 1.0 Pro previous output — revealing default innerwear

Before — With the revealing default innerwear (camisole + short underwear) combination, the upstream safety filter was tightened and generation was sometimes rejected.

Model-Gen 1.0 Pro adjusted output — neutral sports-shorts-based outfit

After — The face, body, style, and pose are kept 100% identical; only the default bottom is adjusted to a neutral sports-shorts-based outfit for stable output.

The model's identity (face, body, style, and pose) is unchanged — only the default bottom outfit is adjusted.

Good to know

  • Due to the nature of the policy, generation may rarely be rejected, but retrying generates it normally.

  • Rejected attempts are not charged any credits. Feel free to retry.

Where to use

Generate a model as usual in Fashion AI's Model-Gen, and it will output with the adjusted default outfit. → Explore Fashion AI

FAQ

Q. Why did the default outfit change? It's an adjustment to ensure stable output in line with the generation engine's strengthened safety policy. The model's identity (face, body, style) is preserved.

Q. Are credits charged if generation is rejected? No. Failed attempts are not charged. Retrying generates it normally in most cases.

Q. Are other Fashion AI apps affected? This adjustment applies only to Model-Gen 1.0 Pro output. Other apps and models are unchanged.


Model-Gen is part of Abocado AI's Fashion AI — one of 7 apps that automate every stage of the fashion workflow.

Fashion AI 7

Explore all 7 apps that automate the fashion workflow at a glance.

Learn more

Model-Gen

The Model Generation app with this update applied. Create a model right now.

Create