Model Context Protocol

DesignerBox MCP turns any AI into a creative studio

Connect DesignerBox to Claude Code, Cursor, ChatGPT and Claude Desktop. Generate on-brand images, product videos and avatars right from your chat, then edit, restyle and reuse your whole design library without leaving the conversation.

Works withClaude DesktopClaude.aiCursorChatGPTMCP Inspector
1

Copy your MCP server URL

You will paste this into your AI client.

https://mcp.designerbox.ai/api
2

Add the connector

In Claude Desktop or Claude.ai, open Settings then Connectors, click Add custom connector, name it DesignerBox and paste the URL.

Open in Claude Desktop

In ChatGPT, open Settings then Connectors, add a connector in developer mode and paste the URL, then authorize the DesignerBox sign-in.

Add DesignerBox to your Cursor MCP config, then authorize when prompted.

{ "mcpServers": { "designerbox": {
  "url": "https://mcp.designerbox.ai/api",
  "transport": "http" } } }

Launch MCP Inspector, choose the Streamable HTTP transport, paste the URL and complete the OAuth prompt.

Generate a personal key in the DesignerBox app under Account then API Keys. It starts with apikey_ and works as a bearer token on every request.

POST https://mcp.designerbox.ai/api
Authorization: Bearer apikey_xxx
Content-Type: application/json
3

Sign in and create

Approve the DesignerBox sign-in once. Your client receives a scoped 30-day token, then you ask it to design.

Start calling tools

No sign-in needed. Your assistant calls the JSON-RPC 2.0 endpoint over HTTPS with your bearer token.

Using Claude Code, a script or CI? Generate an API key from Account then API Keys. Read the docs

Not using Claude? See all connect options

Image generation

Turn a prompt into on-brand images

Ask for product shots, campaign visuals or social posts. Your assistant reads your active brand profile and hands back finished, on-brand images in about 30 seconds.

Generate an image
1
Generate 4 studio product shots of our new running shoe, use our brand profile
Nano Banana Pro1:14 images
AI-generated studio product photography of a running shoe

Here are four on-brand shots. Want a lifestyle set next?

DownloadRegenerate
Product video

Turn product pages into launch videos

Paste a product URL or a short brief. Your assistant pulls the details and returns a polished, ready-to-post launch video with motion and audio.

Create a video
2
Use this product URL to make an 8s launch video: shop.acme.com/sneakers
Veo 3.116:98sAudio
Frame from an AI-generated product launch video0:08

Here is your launch cut. Want a 9:16 version for Reels?

DownloadRegenerate
Virtual try-on

Put your product on any model

Drop in a garment and a model brief. Your assistant dresses the model and returns editorial-ready shots you can drop straight into a campaign.

Try it on
3
Put this jacket on a model in a street-style shoot
Nano Banana Pro3:4
AI virtual try-on of a jacket on a model

Styled and shot. Want more angles?

DownloadMore angles
AI avatars

Create a reusable AI avatar

Describe a face or brand persona once. Reuse the exact same avatar across product shots, ads and social content, just by calling its name.

Create an avatar
4
Create a brand avatar: warm, approachable founder in her 30s
Nano Banana Proportrait
AI-generated avatar portrait

Avatar ready and saved. Reuse it any time by name.

Save avatarRegenerate
Cinematic motion

Animate product shots with cinematic motion

Start from a still, pick a camera move, and your assistant builds a cinematic clip that orbits your product in seconds.

Animate an image
5
Animate this bag photo into a 6s cinematic shot with a slow 360
Seedance 2.016:96s
Cinematic still from an AI product animation0:06

Rendered your orbit. Download it or push it to your gallery.

DownloadPush to gallery
Brand workflows

Keep your whole library on brand

Run a named workflow and your assistant audits recent designs, flags anything that drifts from your brand, and suggests fixes before you approve.

Audit my library
6
Run /audit-gallery and flag designs that drift from our brand
Brand audit8 / 12 on brand
Summer promo 03
Off-palette: uses teal, brand is coral
Founder headshot v2
Tone drift: casual, brand voice is refined
Launch banner A
On brand

Three flags found. Want me to draft brand-profile updates?

View allFix with /brand-sync

The best AI models, in one conversation

Your assistant picks the right engine for the job. You just describe the result you want.

Nano Banana Pro
Google
Image
GPT Image 2
OpenAI
Image
Seedream 5
ByteDance
Image
FLUX 2 Flex
Black Forest Labs
Image
Veo 3.1
Google
Video
Sora 2 Pro
OpenAI
Video
Seedance 2.0
ByteDance
Video
Kling 2.6 Pro
Kuaishou
Video

Every DesignerBox app, one message away

Flat lays, product ads, try-on, retouch and more. Every one-click app is available to your assistant through MCP.

Browse all apps

From concept to content in one conversation

Tell your assistant what you are working on. It picks the models and delivers production-ready results.

Start now

43 tools, grouped the way you work

Read tools work on every plan. Generation tools use AI credits on a paid plan.

Designs & search4 tools
List, fetch and full-text search your designs and folders.
Image & video generation7 tools
Create images in about 30 seconds and submit async video jobs.
Avatars5 tools
Generate and manage reusable AI avatars.
Brand profiles5 tools
Read and patch brand voice, themes, niche and assets.
Assets, packs & templates7 tools
Browse uploads, photo packs, templates and stock photos.
Pipelines & workflows5 tools
Read pipelines, community workflows and skills.
Whiteboards & chats5 tools
Read whiteboards, galleries and AI Assistant history.
Teams, plan & discovery5 tools
Switch teams and read plan, credits and the tool catalog.

5 tools use AI credits. 2 destructive tools, cancel avatar and cancel video, always confirm first.

Auto-loaded resources

Read-only context your assistant pulls at the start of a chat, no permission prompt.

designerbox://account/summaryPlan, AI credits, design count, active team and folders.
designerbox://brand-profiles/activeYour active brand voice, tone, themes and niche.
designerbox://teams/currentWhich team this connection is scoped to.
designerbox://recent-designsYour 10 most recent designs, pre-loaded into context.

Named workflows

Slash-menu shortcuts that orchestrate several tools and confirm before changing anything.

/weekly-design-batchPropose weekly design themes from your brand and recent work.
/audit-galleryFlag recent designs that drift from your brand. Read-only.
/brand-syncSurface stale brand fields and suggest updates you approve.
/regenerate-with-styleRecreate an existing design in a different style.

Pick the plan that matches your usage

Every plan reads designs and browses photo packs. Generation unlocks as you upgrade.

Every plan

Free

Read designs, search photo packs, list folders and brand profiles, and check plan limits. 112 credits, 1 channel, 10 MB storage.

Most popular

Basic, Pro & Premium

AI image, video and avatar generation, brand updates and multi-team workspaces. 500 to 2,500 credits a month.

For teams

Ultra

Everything at scale: 8,000 credits a month, unlimited storage and unlimited channels.

Got any questions left?

Everything you need to connect DesignerBox to your AI assistant.

How does DesignerBox connect to AI assistants?
DesignerBox runs a Model Context Protocol (MCP) server. You add one URL to your AI client, approve an OAuth sign-in once, and the assistant can call DesignerBox tools over Streamable HTTP. No plugin to install.
Which AI clients are supported?
Claude Desktop, Claude.ai, Cursor, ChatGPT and MCP Inspector, plus anything that speaks MCP over Streamable HTTP.
What can I create through MCP?
On-brand images, product and launch videos, reusable avatars, virtual try-on shots and cinematic clips. You can also search your designs, manage brand profiles, and browse photo packs and templates in plain language.
Which AI models can it use?
Nano Banana Pro, GPT Image 2, Seedream 5 and FLUX 2 Flex for images. Veo 3.1, Sora 2 Pro, Seedance 2.0 and Kling 2.6 Pro for video. Your assistant picks the right engine for each job.
Do I need an API key?
Not for AI clients. They use OAuth. For Claude Code, scripts or CI, generate a personal key in the DesignerBox app under Account then API Keys. It starts with apikey_ and works as a bearer token.
How do credits and pricing work?
Read tools work on every plan, including Free. AI generation uses credits on a paid plan: 500 on Basic, 1,000 on Pro, 2,500 on Premium and 8,000 on Ultra each month.
Is it secure?
OAuth gives your client a scoped access token, never your password. Tokens last 30 days and you can revoke them any time. Read tools run freely. Write and generation tools ask for confirmation in the chat first.
How long does generation take?
Images return in about 30 seconds. Video is asynchronous: the assistant submits the job and polls until the file is ready, then hands you the link.
Can it use my existing designs and brand profiles?
Yes. Four resources auto-load your account summary, active brand profile, current team and 10 most recent designs, so the assistant has context without asking each session.
What is the MCP server URL?
Paste the MCP server URL shown above when your client asks for an MCP server, or open the Connect page for one-click setup.

Your design team now has an AI front door

Connect once. Then ask any AI assistant to design, generate and manage your work.