Popcraft MCP
A remote MCP connector that lets AI agents generate video, images, music and sound effects through Popcraft using OAuth sign-in.
At a Glance
100 credits to try, no card. One job at a time, 720p up to 5 seconds, three saved avatars. Outputs are watermarked and for personal use; no monthly credit allowance.
Engagement
Available On
Alternatives
Listed Oct 2026
About Popcraft MCP
Popcraft MCP is a Model Context Protocol connector from Popcraft that lets AI agents such as Claude, ChatGPT, Cursor, Claude Code and Codex create video, image and audio content from within a conversation. It connects through a single endpoint with OAuth sign-in and no API key, and generations use the user's Popcraft credits.
What It Is
Popcraft MCP exposes Popcraft's multi-model generation platform as a set of tools for any MCP client that supports Streamable HTTP with OAuth. The site describes 21 tools covering creation, enhancement, media and reference handling, job tracking, model catalog lookup and account control. An agent takes a brief, previews the credit cost, calls the generation tools and returns result URLs.
How the Workflow Runs
Agents add the endpoint as a custom connector, approve access in a browser, and then request deliverables in chat. Generation is asynchronous: the agent submits a job, receives a job ID, and polls until result URLs are ready. Every generation tool supports a cost preview via get_cost, so the agent can state the exact credit price before submitting anything. Outputs also appear in the user's Popcraft web-app history for further editing on the Canvas, upscaling or download.
Skills and Model Catalog
Agents can load production skills served from Popcraft's server before generating. Two are shown: ugc-ad, which turns a product photo into a vertical creator-style ad via a storyboard, realism pass and cost preview before a Seedance 2.0 clip, and thumbnail-pro, a thumbnail workflow with concept frameworks, an eleven-block prompt and identity lock. The model catalog, explorable through the models_explore tool, lists video models such as Seedance 2.0, Gemini Omni 1.1 Flash and Kling 3, and image models such as GPT Image 2, Seedream 5 Pro, Nano Banana Pro and Midjourney.
Media References and Account Tools
Users can import any public image, video or audio URL as a reference for style, start frames or image-to-video sources, and chain generated results into later generations. Account tools report credit balance, plans and transactions. Popcraft says the connector works on the free tier, with some premium models reserved for paid plans.
Community Discussions
Be the first to start a conversation about Popcraft MCP
Share your experience with Popcraft MCP, ask questions, or help others learn from your insights.
Pricing
Free
100 credits to try, no card. One job at a time, 720p up to 5 seconds, three saved avatars. Outputs are watermarked and for personal use; no monthly credit allowance.
- 100 credits to try
- 1 concurrent task
- 720p up to 5 seconds
- 3 saved avatars
- Watermarked outputs, personal non-commercial use
Starter
For a first look at AI video
- 1,000 credits/mo
- 2 concurrent
- All video & image models
- Commercial license
- 2 voice clones
- Watermark-free downloads
Plus
For occasional video work
- 4,000 credits/mo
- 4 concurrent
- Commercial license
- 5 voice clones
- Unlimited Nano Banana / Seedream image models
- Watermark-free downloads
Pro
For creating most days
- 12,000 credits/mo
- 8 concurrent
- Seedance credits -40% annual / -20% monthly
- Unlimited Seedance 2.0 Fast & Mini daily allowance
- 10% off top-ups
- 15 voice clones
- Watermark-free downloads
Max
For production volume
- 28,000 credits/mo
- 10 concurrent
- Seedance credits -60% annual / -40% monthly
- Unlimited Seedance 2.0 Fast & Mini daily allowance
- 10% off top-ups
- 50 voice clones
- Watermark-free downloads
Enterprise
Custom credit allocation, higher concurrency, dedicated account manager, priority queue
- Unlimited membership
- Custom credit allocation
- Higher custom concurrency
- Access any unlimited models you need
- Priority queue for faster task processing
Capabilities
Key Features
- Single endpoint with OAuth 2.1 sign-in, no API key
- Text-to-video, image-to-video and first/last-frame video generation
- Image generation and editing across multiple models
- Audio generation: TTS, sound effects and music
- Image and video upscaling
- Reference import from public URLs and signed uploads
- Cost preview before every generation
- Asynchronous job polling with result URLs
- Production skills served live (UGC ad, thumbnails)
- Model catalog with capabilities and credit costs
- Generation history and reusable characters and styles
- Credit balance, plans and transaction tools
