Purpose
Movie Scene Map MCP server provides AI assistants with structured data about film and television filming locations and the real-world settings of fictional works. It connects to the Movie Scene Map atlas (called FilmMap before September 1, 2026), enabling AI assistants to answer questions such as where a movie was filmed, which works a location appears in, and which real-world place a game or anime is set in.
Main Capabilities
- Filming location queries for film and television works: movies and TV series are located to real filming locations via Wikidata's filming location property (P915).
- Setting location queries for games, anime, and manga: these works have no filming locations, so they are located to the real-world places where their stories are set via the narrative location property (P840). Each record indicates the type of relationship used.
- Location details: including coordinates, popularity ranking, the list of works produced at that location, and the sources behind each statement.
- Statistics: real-time counts are available via get_statistics; the specific numbers are updated with each release and can also be viewed at https://moviescenemap.com . You can also view them at
- Evidence distinction: most links come from Wikidata statements; some come from locations mentioned in the "filming" section of an English Wikipedia article about a work, or from setting categories archived by editors. These appear in separate fields and indicate the exact article and section read. When citing Wikipedia sources, state "according to Wikipedia".
How to Use
The service is already running online and requires no installation, registration, or API key. Clients can point directly to the hosted endpoint:https://moviescenemap.com/mcp。旧地址 https://thefilmmap.com/mcp It can still respond to clients configured before the rename. The protocol is streaming HTTP, stateless, POST-only (GET returning 405 is normal).
To run it yourself, Docker can also be used, with a public copy of the atlas already built into the image. HOST defaults to 127.0.0.1; in containers it is set to 0.0.0.0 because the default value is unreachable outside the container. The server reads a single file, atlas-compact.json, which is the website's own transport format and is rebuilt on each deployment.
Use Cases
Suitable for scenarios that need to integrate information about film and television filming locations or the settings of fictional works into AI assistant answers, such as travel planning, film and television research, and content creation assistance. Because the data comes from open knowledge bases, it can also be used for data analysis and visualization.
Limitations and Notes
- Data coverage is uneven: some works are recorded only at the country level, appear as countries, and have no coordinates. Works that cannot be located at all are published in the work list at moviescenemap.com/gaps/ so they can be fixed on Wikidata.
- Location is not the same as scene: P915 only records that a work was filmed at a location; which specific scene was filmed where is not in the data and cannot be inferred from the data.
- The two types of evidence cannot be merged: Wikidata statements and Wikipedia sources must be stated separately.
- The server code is licensed under MIT, and the dataset is licensed under CC0; photos and Wikipedia text referenced via URLs retain their own licenses.
- Regarding charges, prices, registration requirements, etc., no information is provided in the materials.
