GitHub hosts HunyuanVideo, Tencent's open-source framework for large-scale video generation models, enabling AI-driven video creation.
Intelligence API
€9/mStarting price
0Popularity
About Intelligence API
Thisispaper Intelligence API provides programmatic access to a curated catalog of over 2,500 architecture, design, photography and art projects selected by human editors since 2011. The catalog exposes search, similarity and taste metadata through both a RESTful HTTP API and a native Model Context Protocol (MCP) server for AI agents. The MCP server allows direct integration with clients like Claude, ChatGPT, Codex and Cursor without authentication, though anonymous access is limited to five results per call. A free API key grants 200 calls per month with full result sets and higher limits. The REST endpoints support lexical search across multiple fields and return structured JSON responses with rate-limit headers. Taste-based search enables filtering by curated dimensions such as material, light, color, mood and composition, providing editorial nuance not typically found in automated systems. The API also includes tools for browsing by recency, discovering random projects, retrieving full project metadata and generating API keys directly within agent conversations.
Key features
- Search projects by keywords
- Filter by curated taste dimensions
- Browse newest projects
- Discover random projects
- Find similar projects via visual and semantic neighbors
- Retrieve full project metadata
- Generate API keys mid-conversation
- REST and MCP endpoints
Use cases
- Finding reference projects for design inspiration
- Researching architectural styles and materials
- Generating curated content for art and design publications
Pros
- Curated taste metadata written by human editors
- Dual access via REST HTTP and native MCP server
- Anonymous preview mode without signup
- Free tier with 200 calls per month
- Supports taste-based search beyond keyword matching
Cons
- Anonymous access limited to 5 results
- Free tier capped at 200 calls per month
- Requires API key for full REST access
- No bulk export or download functionality
Frequently asked questions about Intelligence API
What is the Intelligence API?
The Intelligence API provides programmatic access to a curated catalog of over 2,500 architecture, design, photography, and art projects selected by human editors since 2011. It exposes search, similarity, and taste metadata through both a RESTful HTTP API and a native Model Context Protocol (MCP) server for AI agents.
Who is the Intelligence API for?
The API is designed for developers, researchers, and AI agents needing high-quality, editorially curated visual and design content. It suits users in fields like architecture, design, photography, and art who require nuanced, taste-based search capabilities.
How does the MCP server work with AI agents?
The MCP server allows AI agents like Claude, ChatGPT, Codex, and Cursor to query the catalog natively without authentication, though anonymous access is limited to five results per call. A free API key grants 200 calls per month with full result sets.
What can I do with the REST API?
The REST API supports lexical search across multiple fields and returns structured JSON responses with rate-limit headers. It requires an API key for access, which can be self-served, and all endpoints are CORS-open for browser apps.
How do I get started with the Intelligence API?
To get started, you can either use the MCP server by adding the server URL to your AI client (e.g., Claude, ChatGPT) or make REST API calls using a self-served API key. The MCP server can be added via settings in supported clients, while REST calls require passing the API key as a bearer token.
What are the unique features of the taste-based search?
Taste-based search allows filtering by curated dimensions such as material, light, color, mood, composition, and subject, which were written during human curation. This provides editorial nuance not typically found in automated systems and ranks results by how many dimensions match.
