The MCP connector

The MCP connector lets an AI assistant use your Voplica account inside a chat: it finds models and styles, starts image and video runs on your account, follows them and shows what they made. It is added to Claude as a custom connector, by its URL. It is not the API: it signs in with your Voplica account, and it needs no API key.

Add it to Claude

Open the studio’s MCP page, voplica.com/mcp, and choose Connect Claude to Voplica. It opens Claude’s add-connector dialog, already filled in with the connector’s URL: add the connector there. Or add a custom connector in Claude yourself, with this URL:
```
https://mcp.voplica.com/mcp
```
Claude opens Voplica’s own sign-in page. Sign in with your Voplica account and, if asked, grant access. No API key is needed.

Claude Code can add the connector by the same URL, and it signs in the same way, with your Voplica account.

When ChatGPT can be connected, the studio’s MCP page shows how; in ChatGPT, Voplica shows content up to PG-13.

What it can do

The connector has these tools. Claude asks you before it calls one that costs credits.

Tools	What they do
`get_account`, `list_models`, `list_styles`, `list_characters`, `list_character_looks`, `list_my_content`, `list_my_runs`, `get_run`, `get_content`, `estimate_cost`	Read your account, the models and styles, your characters and their looks, your content and your runs; follow a run; estimate what a run would cost.
`generate_image`, `edit_image`, `enhance_image`	Start image runs: generate, edit with an instruction, upscale or refine faces and hands.
`generate_video`	Start a video run.
`cancel_run`	Stop a run that is still going.
`play_video`	Play a finished run’s video in the chat, in the Voplica player.
`get_video_link`	Used by the player itself, to load the video.

Every run the connector starts is an ordinary Voplica run: it appears in your history in the studio.

What a chat shows

What it makes is private. Every run the connector starts, and everything it makes, is private to you.
Up to R in a chat. An image is shown inline only up to R, or lower if your own settings ask for less. X and XXX are never shown in a chat, inline or as a download: open them in the studio.
An unrated item is rated first. An item with no rating yet gets a quick estimate before it is shown; if it can’t be rated, it isn’t shown in the chat.
The same content rules. The connector’s runs follow the Content policy, as every run does.

How it is paid

Runs started from the connector are paid like studio runs: from your free credits, then plan credits, then Anytime credits, then Earned credits (a third-party model never takes free credits). That is unlike the API, which uses Anytime and Earned credits only (see Billing and limits). Every answer to a paid call states the run’s estimated cost and your balance.

Each account also has limits on the connector’s paid calls: a number of calls per ten minutes and an estimated amount of credits per day. A call over a limit is refused with the reason, and the same call sent again within ten minutes returns the run it already started instead of starting another.