Turn PDFs, Office files, a website or a brief into narrated videos with captions. Pay per video.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Bearer <your ABContents API key>. Only needed for create_video and get_balance; the other tools are free.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Make a video (vertical 9:16 by default; also 4:5, 1:1, 16:9) with voice, captions and music from a brief, files (PDF, Word, Excel, PowerPoint, CSV, text, photos) and/or a public website (site_url: we read its text and can use its logo). Spends tokens from the API key in the Authorization header. Returns the order; poll get_video. The script uses only facts from the brief and files.
id from list_styles
What the video is about. Required if no files.
Optional: we email the order link when the video starts and the download link when it is ready.
Video size: 9:16 vertical (default), 4:5, 1:1 or 16:9.
A public website to make the video from (its text, and its logo if site_logo is auto). Use instead of or with files/brief.
auto: use the best logo found on site_url when no logo file is sent.
How the narrator should say the brand, spelled as it sounds (e.g. "Menunoud" for MenuNode in Spanish). Captions keep brand_name.
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http abcontents https://abcontents.com/api/mcp
Headers
Authorization (optional, not in the config above): Bearer <your ABContents API key>. Only needed for create_video and get_balance; the other tools are free.Make a video (vertical 9:16 by default; also 4:5, 1:1, 16:9) with voice, captions and music from a brief, files (PDF, Word, Excel, PowerPoint, CSV, text, photos) and/or a public website (site_url: we read its text and can use its logo). Spends tokens from the API key in the Authorization header. Returns the order; poll get_video. The script uses only facts from the brief and files.
| Argument | Type | Description |
|---|---|---|
| kind* | string | |
| logo | any | |
| brief | string | What the video is about. Required if no files. |
| string | Optional: we email the order link when the video starts and the download link when it is ready. | |
| files | array | |
| music | string | |
| style* | string | id from list_styles |
| voice | string | |
| format | string | Video size: 9:16 vertical (default), 4:5, 1:1 or 16:9. |
| language* | string | |
| site_url | string | A public website to make the video from (its text, and its logo if site_logo is auto). Use instead of or with files/brief. |
| site_logo | string | auto: use the best logo found on site_url when no logo file is sent. |
| brand_name | string | |
| music_file | any | |
| duration_seconds* | integer | |
| brand_pronunciation | string | How the narrator should say the brand, spelled as it sounds (e.g. "Menunoud" for MenuNode in Spanish). Captions keep brand_name. |
{
"type": "object",
"$defs": {
"file": {
"type": "object",
"required": [
"name",
"content_base64"
],
"properties": {
"name": {
"type": "string"
},
"content_base64": {
"type": "string"
}
},
"additionalProperties": false
}
},
"required": [
"kind",
"language",
"style",
"duration_seconds"
],
"properties": {
"kind": {
"enum": [
"marketing",
"topic"
],
"type": "string"
},
"logo": {
"$ref": "#/$defs/file"
},
"brief": {
"type": "string",
"maxLength": 600,
"description": "What the video is about. Required if no files."
},
"email": {
"type": "string",
"maxLength": 254,
"description": "Optional: we email the order link when the video starts and the download link when it is ready."
},
"files": {
"type": "array",
"items": {
"$ref": "#/$defs/file"
},
"maxItems": 9
},
"music": {
"enum": [
"auto",
"ambient",
"chip",
"cine",
"funk",
"lofi",
"lux",
"pop",
"quirky",
"synth",
"tension",
"own",
"none"
],
"type": "string"
},
"style": {
"type": "string",
"description": "id from list_styles"
},
"voice": {
"enum": [
"f",
"f2",
"f3",
"f4",
"f5",
"m",
"m2",
"m3",
"m4",
"m5"
],
"type": "string"
},
"format": {
"enum": [
"9:16",
"4:5",
"1:1",
"16:9"
],
"type": "string",
"description": "Video size: 9:16 vertical (default), 4:5, 1:1 or 16:9."
},
"language": {
"enum": [
"en",
"es"
],
"type": "string"
},
"site_url": {
"type": "string",
"maxLength": 500,
"description": "A public website to make the video from (its text, and its logo if site_logo is auto). Use instead of or with files/brief."
},
"site_logo": {
"enum": [
"auto",
"none"
],
"type": "string",
"description": "auto: use the best logo found on site_url when no logo file is sent."
},
"brand_name": {
"type": "string",
"maxLength": 60
},
"music_file": {
"$ref": "#/$defs/file"
},
"duration_seconds": {
"type": "integer",
"maximum": 300,
"minimum": 30
},
"brand_pronunciation": {
"type": "string",
"maxLength": 80,
"description": "How the narrator should say the brand, spelled as it sounds (e.g. \"Menunoud\" for MenuNode in Spanish). Captions keep brand_name."
}
},
"additionalProperties": false
}Tokens left on the API key sent in the Authorization header.
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Price in tokens (1 token = USD 1) and estimated wait, queue included. Free.
| Argument | Type | Description |
|---|---|---|
| style* | string | |
| duration_seconds* | integer |
{
"type": "object",
"required": [
"style",
"duration_seconds"
],
"properties": {
"style": {
"type": "string"
},
"duration_seconds": {
"type": "integer",
"maximum": 300,
"minimum": 30
}
},
"additionalProperties": false
}Status of a video: step, progress, seconds left, and the download URL when ready.
| Argument | Type | Description |
|---|---|---|
| id* | string |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string"
}
},
"additionalProperties": false
}Background music choices with preview URLs ("auto" lets the script writer pick, "none" is voice only). Free.
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}The 20 video styles (id, name, 2D/3D). Free.
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Narrators for a language, with a sample mp3 URL each. Free.
| Argument | Type | Description |
|---|---|---|
| language* | string |
{
"type": "object",
"required": [
"language"
],
"properties": {
"language": {
"enum": [
"en",
"es"
],
"type": "string"
}
},
"additionalProperties": false
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 20, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.