AI & Machine Learning

imgica.ai

An MCP server for AI media generation that lists available models, voices, and filter styles and estimates token costs for generation requests.

1 endpoint26 known toolsFirst detected October 4, 2026Last detected October 4, 2026

ENDPOINT 1

https://imgica.ai/mcp

No auth detected

MCP server metadata

Name: imgica
Version: 1.0.0

Capabilities

tools.listChangedcompletionsresources.listChangedprompts.listChanged

Server instructions

Imgica turns prompts into images, video, music, voice, vectors and 3D on the user's prepaid token balance. Start with list_models (no sign-in needed) and get_model for a model's exact parameters and price, then estimate_cost before a paid call. Every create tool (generate_image, generate_video, edit_image, ...) charges tokens, submits the job, waits for it, and returns the output URLs in the same call. If a create tool answers that the job is still running (video and music can take minutes), call wait_for_job right away with the kind and jobId it gives you, and keep calling it until the job completes. Do not ask the user to check back later, and never tell the user a result is ready before you hold its URL. Outputs are URLs on Imgica's storage. Pass an image URL from one job as the input of another (edit_image, upscale_media, generate_video with image). For a file the user has locally, call upload_media with base64 or a public URL first; it returns an input URL every tool accepts. A failed job refunds its tokens automatically. Use idempotencyKey on create tools when retrying so a retry never charges twice.

Known tools 26

list_models

List the models or engines available for a kind of work, with prices in tokens.

Inferred read-only

get_model

One model with its full parameters: sizes or ratios, durations and resolutions, input limits, per-mode forms (video) and the exact token prices.

Inferred read-only

estimate_cost

The exact tokens a create tool would charge for the given arguments, computed by the same pricing the charge uses.

Potential side effects

list_voices

Voices for text_to_speech.

Inferred read-only

list_filter_styles

The style presets apply_filter accepts (anime, watercolor, cyberpunk, ...), with preview images.

Inferred read-only

get_account

The connected user's plan and token balance: tokens left, the protected top-up portion, plan expiry, how many jobs may run at once and whether premium models are available.

Inferred read-only

generate_image

Create one or more images from a prompt with an Imgica image model (list_models kind "image").

Potential side effects

edit_image

Edit one or more images with the Editing Studio: features ai-edit, change-region, magic-eraser, crop-expand, change-light, camera-angle, color-palette, style-transfer, upscale, remove-background.

Potential side effects

blend_images

Combine several reference images with a prompt into a new image (the Image Reference studio).

Inferred read-only

enhance_prompt

Rewrite a rough prompt into a detailed, cinematic one, the way the studio's magic prompt does.

Inferred read-only

generate_video

Run any Video Studio model: text-to-video, image-to-video, extend, edit, upscale, lip sync, avatars, face swap, background removal (list_models kind "video"; get_model shows each mode's form).

Potential side effects

upscale_media

Upscale an image to 2k, 4k or 8k, or a video to 1080p, 2k or 4k (billed per second of the clip, 5 to 60 seconds).

Inferred read-only

remove_background

Cut the background out of an image (engines standard, ideogram, bria, or cutout which takes a prompt naming the object to keep) or a video (billed per second, 5 to 120 seconds; video inputs must be Imgica-issued URLs).

Inferred read-only

generate_vector

Create an SVG from a prompt (engines budget, standard, pro; ratios 1:1, 4:3, 3:4, 16:9, 9:16).

Potential side effects

apply_filter

Restyle a photo with a preset (list_filter_styles) or your own prompt, at fast, standard or fine quality.

Inferred read-only

swap_face

Put the face from one image onto another (engines standard, pro, head-swap, infinite-you).

Inferred read-only

generate_3d

Make a 3D model (GLB) from a prompt or an image (engines rapid, standard, pro).

Inferred read-only

generate_music

Make a song with sung lyrics (mode song: engines rapid, standard, cinematic, pro; lyrics required) or an instrumental (mode bgm: engines rapid and pro price by durationSeconds 10 to 180, standard is a fixed length).

Inferred read-only

text_to_speech

Narrate text (up to 10,000 characters, 5,000 on pro) in a chosen voice: engines rapid (MiniMax voices), standard and pro (ElevenLabs voices, see list_voices).

Inferred read-only

get_job

The current state of a job: status (created, processing, completed, failed), its output URLs when done, the error and whether tokens were refunded when failed.

Inferred read-only

wait_for_job

Long-poll a job until it completes or fails, up to timeoutSeconds (default 45, max 240).

Inferred read-only

list_jobs

The user's jobs of one kind, newest first, with outputs for completed ones.

Inferred read-only

delete_job

Remove a job from the user's library.

Potential side effects

upload_media

Bring a file in so the other tools can use it: give a public URL (the server fetches it) or base64 data with its mimeType.

Inferred read-only

search

Search the user's Imgica library (completed images, videos, music, voice, vectors, 3D) and the model catalog.

Inferred read-only

fetch

Fetch the full record behind a search result id: a job with its outputs, or a model with its parameters and prices.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.imgica]
url = "https://imgica.ai/mcp"
enabled = true

Claude Code

.mcp.json

{
  "mcpServers": {
    "imgica": {
      "type": "http",
      "url": "https://imgica.ai/mcp"
    }
  }
}

Claude Desktop

Settings → Connectors → Add custom connector

Name: imgica
Remote MCP URL: https://imgica.ai/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "imgica": {
      "url": "https://imgica.ai/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

{
  "servers": {
    "imgica": {
      "type": "http",
      "url": "https://imgica.ai/mcp"
    }
  }
}

Generic MCP

Client-specific MCP configuration

{
  "name": "imgica",
  "transport": "streamable-http",
  "url": "https://imgica.ai/mcp"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.