Stability AI MCP server for your AI

Ask for a picture, a cleaner product shot or a backing track, and see it in the chat with a link to download it. Your AI picks the right Stability AI model and waits for longer jobs to finish.

PopMCPStability AI

Check my credits, then make a wide image of a red fox in the snow

Contact sales Connect

Supported Stability AI tools

Every endpoint in Stability AI's REST API, with tools that take a link or an earlier result as their input, check the settings before any credit is spent, and hand back each finished file as a download link.

37 tools in the full catalog: 7 read and 30 write. 33 load by default, and the rest are found with tool search.

stability_erase_object

Remove something from an image, like a blemish or a stray item, and fill the space naturally.

stability_generate_audio

Make music or sound from a text prompt with Stable Audio 3.0, up to about six minutes.

stability_generate_audio_v2

Make music or sound from a text prompt with Stable Audio 2, up to 190 seconds, answering right away.

stability_get_audio_result

Fetch finished audio by generation id, from stability_generate_audio, stability_transform_audio or stability_inpaint_audio.

stability_inpaint_audio

Replace one section of an audio file, using Stable Audio 3.0.

stability_inpaint_audio_v2

Replace one section of an audio file, using Stable Audio 2.5, answering right away.

stability_inpaint_image

Replace a marked part of an image with what a prompt describes.

stability_outpaint_image

Extend an image outward, like making a portrait wider.

stability_remove_background

Cut out the subject of an image and make the background transparent.

stability_search_and_recolor

Change the colour of an object by naming it, with no mask.

stability_search_and_replace

Swap an object in an image by naming it, with no mask.

stability_start_replace_background_and_relight

Put a subject on a new background and match the lighting, for product and portrait shots.

stability_transform_audio

Change an audio sample with a text prompt, using Stable Audio 3.0.

stability_transform_audio_v2

Change an audio sample with a text prompt, using Stable Audio 2, answering right away.

stability_control_sketch

Turn a sketch or line drawing into a finished image.

stability_control_structure

Make a new image that keeps the layout of an input image, like rebuilding a scene or a character pose.

stability_control_style

Make a new image from a prompt in the style of a reference image.

stability_start_creative_upscale

Rebuild a small or badly degraded image at up to 4K, inventing detail guided by a prompt.

stability_style_transfer

Repaint one image in the style of another.

stability_upscale_image_conservative

Upscale an image up to 4K while keeping it as it is.

stability_upscale_image_fast

Make a small image four times larger, cheaply, without changing it.

stability_generate_image_core

Make an image from a text prompt quickly and cheaply with Stable Image Core.

stability_generate_image_sd3

Make an image with a Stable Diffusion 3.5 model, from text or from a starting image.

stability_generate_image_ultra

Make the highest quality image from a text prompt with Stable Image Ultra.

stability_get_account

Get the Stability account this API key belongs to: its id, email and organizations.

stability_get_balance

Check how many credits are left, before spending them or after an error says credits ran out.

stability_list_engines

List the v1 engines this key can use with the stability_sdxl_ tools: id, name, type and description.

stability_sdxl_image_masking

Change only part of an image with a prompt using SDXL 1.0.

stability_sdxl_image_to_image

Change an image with a prompt using SDXL 1.0.

stability_sdxl_text_to_image

Make images from text with the older SDXL 1.0 model, when someone asks for it or needs up to ten images in one call.

stability_generate_3d_fast

Make a textured 3D model from one image, quickly.

stability_generate_3d_point_aware

Make a 3D model from one image with Stable Point Aware 3D, which builds the unseen back of the object well.

stability_get_result

Fetch a finished image by generation id, from stability_start_creative_upscale or stability_start_replace_background_and_relight.

stability_get_creative_upscale_result

Fetch a creative upscale by id, from stability_start_creative_upscale.

stability_legacy_get_upscale_result

Fetch a result of stability_legacy_start_upscale by id.

stability_legacy_inpaint_image

The older v2alpha inpaint route.

stability_legacy_start_upscale

The older v2alpha creative upscale route, which spends credits.

Connect Stability AI to Claude, ChatGPT and Claude Code

Credentials off one Stability AI screen, then an endpoint your client signs into. Nothing to deploy, and nothing to install.

Paste an API key from platform.stability.ai/dashboard/api-keys. PopMCP checks it by reading your account and credit balance, which costs nothing. A key belongs to one account, so there is nothing else to choose.

API key••••••••••••3f7a

ConnectionVerified

Drop the PopMCP URL into Claude, ChatGPT, Claude Code, Codex or any other supported client. It signs in through PopMCP with OAuth, so your Stability AI credential never leaves the server.

app.popmcp.com/mcp

Claude wants to connectApprove

Ask for something harmless before the agent changes anything. Writes run as soon as it calls them, so confirm it reads the right account first.

Check my credits, then make a wide image of a red fox in the snow

stability_get_balanceok

Then just ask for the work

“Check my credits, then make a wide image of a red fox in the snow”
- get_balance
- generate_image_core
“Remove the background from this product photo and give me a transparent PNG”
- remove_background
“In this room photo, replace the sofa with a green velvet one”
- search_and_replace
“Put this product on a marble kitchen counter with soft light from the left”
- start_replace_background_and_relight
- get_result
“Upscale the image you just made so I can print it”
- upscale_image_conservative
“Make a one minute lo-fi track with soft piano for my video”
- generate_audio
- get_audio_result
“Turn this photo of a chair into a 3D model I can download”
- generate_3d_fast

Contact support View docs

Use Stability AI from the assistant you already open

One hosted endpoint, registered once per client. The same tools and the same permissions, whichever window the request comes from.

Claude DesktopCustom connector

Add the PopMCP URL under Customize → Connectors and the tools appear in every chat.

Claude WebCustom connector

The same endpoint, registered in the browser. Nothing to install on the machine you are working from.

ChatGPTCustom connector

CodexDesktop & CLI

Wire the endpoint into Codex or the Codex CLI when a task needs live business context, not a fixture.

OpenCodeCoding agent

ZCodeDesktop app

Add PopMCP as an MCP server in Z.ai's desktop app and work the account from the same window.

Contact support View docs

FAQ

Stability AI MCP server questions

What teams ask before putting Stability AI behind an AI client.

What does the Stability AI MCP server need to connect?

A Stability AI API key from platform.stability.ai/dashboard/api-keys. Paste it into PopMCP. There is no app to review and nothing to deploy.

Will my AI spend my Stability AI credits?

Only if you let it. Each image, song or 3D model spends credits, and failed generations are free. New PopMCP connections start read-only, so your AI can check your balance but cannot make anything until an owner or admin allows write actions. Each tool tells your AI what it costs.

How do I get the files my AI makes?

Images appear in the chat when your client can show them and they are small enough. Every image, audio file and 3D model also comes with a download link that lasts one hour, so save what you want to keep.

Can my AI use my own images and audio?

Yes. Give it a public link, or a download link from an earlier result, up to 10 MiB. A file path on your computer will not work, because PopMCP runs on a server. So your AI can make an image, then upscale or edit that same image in the next step.

What about jobs that take longer?

Creative upscaling, background replacement and Stable Audio 3.0 run as jobs. PopMCP waits for them for a while, and if one is still going your AI gets its id and can fetch the result later, so nothing is paid for twice. Stability AI keeps those results for 24 hours.

Which models does it cover?

Stable Image Ultra and Core, Stable Diffusion 3.5, SDXL 1.0, the upscalers, edit and control tools, Stable Audio 2 and 3.0, Stable Fast 3D and Stable Point Aware 3D. Stable Video is not included, because Stability AI removed it from its API.

Do I have to run or host the Stability AI MCP server myself?

No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.

Which AI clients can use this MCP server?

Any client that supports a remote Streamable HTTP MCP server: Claude Desktop, Claude Web, Claude Code, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.

How do I connect Stability AI to Claude Code?

Connect Stability AI in PopMCP first, then run claude mcp add --transport http popmcp https://app.popmcp.com/mcp in a terminal. Inside Claude Code, run /mcp and sign in to PopMCP in the browser window it opens, ticking Stability AI on the consent screen. The Stability AI tools are then available in that session, and no Stability AI credential is stored on your machine.

Why do only 33 of the 37 tools load by default?

Loading 37 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 33-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. Every tool stays searchable, and on Scale the owner can load the full catalog for a teammate's connection.

What is the difference between a tool and a trigger?

A tool is something your AI calls, and there are 37 here. A trigger is an event PopMCP watches for on its own. Stability AI has no triggers yet, so everything on this page happens when your AI calls a tool.

Connect Stability AI to your AI in minutes

Launch a hosted MCP endpoint for Stability AI, keep tool access under control, and start testing real workflow prompts right away.

Connect your first MCP