MCP App Store
by KeenEthics
On-Model icon

On-Model MCP Server for ChatGPT

By PiktID

Overview

On-Model is the AI infrastructure fashion brands and retailers use to produce catalog imagery. Connect it and ask in plain language: ChatGPT uploads your product photos, picks a model and runs the job in your On-Model account. - Flat-to-model: dress a model in flat-lay or packshot garments, in your choice of pose, background and size. - Packshots: ghost mannequin, flat lay, white cutout or marketing scenes from raw product photos, even phone shots. - Garment recolor: new colourways across a set of photos, everything else untouched. - Model swap: re-cast existing product photos with a different model, keeping the clothes and background identical. - At scale: saved templates for your house style, bulk runs across tagged products with a preview before anything runs, and automatic quality checks with AI retouching. Results appear in your On-Model account like any other job, ready to review and download.

How to connect

  1. ChatGPT: Open the On-Model listing in the ChatGPT app directory and choose Connect. ChatGPT listing On-Model exposes 31 tools here.

Tools

add_identity

ChatGPT
Create a digital twin of a brand model from one frontal photo, for consistent faces across the catalog. Only use photos of people who have consented to being used as a model. Returns the identity_code. Costs 50 credits per identity (like keeping a generated one): confirm first.

add_preset_to_template

ChatGPT
Add a preset to a template, or remove it. A preset can be in several templates.

bulk_create

ChatGPT
Run the same settings across a whole collection: one job per product tag (up to 100). With dry_run=true (default) nothing is generated and the response shows eligible products, blockers and estimated credits.

create_identity

ChatGPT
Generate NEW synthetic model identities from a text brief (no photo needed). Returns a job_id; review the drafts with get_job (each result carries an image_result_id), then promote_identity the ones to keep. Framing, lighting and outfit are standardized by the platform (clean front-facing portrait). Drafts cost credits per output like other jobs; promotion costs 50 each.

create_packshot

ChatGPT
Generate product-only packshots (ghost mannequin, flat lay, white cutout, marketing scene) from raw garment photos. No model needed. Costs credits: 3/5/10 per output at 1K/2K/4K.

create_preset

ChatGPT
Save a preset (one reusable scene). Give settings directly (free), or let On-Model write them from a text description or a reference image (extraction uses credits). The type cannot change later.

create_template

ChatGPT
Create an empty template. Then add presets with create_preset (template_ids) or add_preset_to_template. Free.

create_template_from_images

ChatGPT
Build a template in a brand's house style from 2-6 of its photos: one preset per image, with the traits the images share worded identically. Saves the template and presets. Costs 1 credit per image.

duplicate_template

ChatGPT
Make your own editable copy of a template and its presets (max 50), e.g. to customise a built-in template. Free.

find_assets

ChatGPT
Find images already in the On-Model account (uploaded here or in the app): by asset ID, text or product tag. The returned file_id goes straight into any generation tool, no re-upload needed.

fix_images

ChatGPT
Ask On-Model's AI retoucher to fix images its reviewer flagged (logos, text, small garment details, artifacts). Pro/Enterprise plans. Costs 1 credit per image actually fixed; check get_job afterwards.

flat_to_model

ChatGPT
Dress a model in flat-lay / packshot garments (virtual try-on). N input garments are combined into one outfit; each shot produces one image. Optional finishing looks require post_process=true. Use look.name='none' when garment color accuracy matters most. Costs credits: 3/5/10 per output at 1K/2K/4K.

garment_recolor

ChatGPT
Create a new colourway: recolor one garment consistently across a set of product images, leaving everything else untouched. Provide at least one of target_color, reference_file_id or instruction. Costs credits: 3/5/10 per output at low/medium/high processing quality.

get_credits

ChatGPT
Show the account's plan and credit balance. Check this before large jobs; generation costs 3/5/10 credits per output at 1K/2K/4K. On Enterprise the balance does not limit or pay for jobs.

get_identity

ChatGPT
Show one model identity by its code (e.g. from get_job's inputs or the app): name, appearance, whether it's On-Model's or the account's own, readiness and a thumbnail.

get_job

ChatGPT
Get a job's status and, when finished, its result images: full-size URLs plus inline thumbnails for quality review.

get_template

ChatGPT
Show a template and its presets, and whether you can edit it.

identity_status

ChatGPT
Preprocessing status of an identity (pending, processing, completed, failed).

image_playground

ChatGPT
Generate or edit an image, with up to four variations. No project or identity is required. Supply a prompt, uploaded images, or saved identities. For a touch-up of a flat-to-model/packshot result, upload its URL first. Costs 3/5/10 credits per variation at 1K/2K/4K; failed variations aren't charged. Returns immediately: use get_job and inspect per-image statuses, not the conversation's top-level status. Continue editing with conversation_id.

list_identities

ChatGPT
List the AI models (identities) available for flat_to_model, model_swap and bulk_create: On-Model's library plus the account's own. Returns a compact summary per model; use filters to narrow the choice.

list_presets

ChatGPT
List saved presets (single scenes: pose, background, lighting, camera, output settings), newest first. Accounts often have hundreds: narrow with search, type, ownership or template_id. Use list_templates for sets of presets.

list_tags

ChatGPT
List product tags (each tag groups the photos of one product). Use the tag ids with bulk_create.

list_templates

ChatGPT
List templates: named sets of presets (scenes) that can be run in one go. Pass a template's id as template_id to flat_to_model, create_packshot or bulk_create to generate one image per preset.

model_swap

ChatGPT
Re-cast existing product photos with a different model. The whole image is regenerated; set match_original_pixels=true to keep clothes and background identical to the photo. Costs credits: 5/7/10 per output at low/medium/high processing quality.

promote_identity

ChatGPT
Turn one generated draft into a permanent identity usable in flat_to_model, model_swap and bulk_create. Costs 50 credits and one identity slot. The identity is private until shared; check readiness with identity_status.

save_template

ChatGPT
Save shots you have tested (flat_to_model / create_packshot shots) as a template: one preset per shot, same fields. Typical loop: write shots, run them, look at get_job, adjust, then save. To refine a saved template: read it with get_template (its presets' settings are shots), change what's needed, re-test with template_presets, then call this with template_id. Free. Read-only templates (e.g. built-in ones) can't be changed: duplicate_template first.

share_identity

ChatGPT
Share an identity with (or remove it from) one of the account's groups.

update_preset

ChatGPT
Edit a preset you own: name, description, settings or processing options. Built-in presets can't be edited; duplicate their template first.

update_template

ChatGPT
Rename a template or change its description or cover image. Fields you don't pass are kept (the API would otherwise clear them).

upload_images

ChatGPT
Upload garment/product photos to On-Model. Returns file_ids and the project_id to pass to the generation tools. Accepts public https URLs (preferred) or base64, up to 20 MB each. Free; starts no job. For a collection, call once per product with a distinct product_tag.

Capabilities

WritesInteractive

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/on-model"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/on-model)