Every collection tool returns 5 items per page with a `start` parameter and totalCount / hasMore / nextStart metadata, so browsing a chart costs several small calls instead of one payload the model has to wade through. Responses are cached in memory for 10 minutes by default, which keeps repeat questions off your RapidAPI quota — that cache is lost when the server restarts.
A Python server over the IMDb API on RapidAPI, covering title search and details, cast and crew, the IMDb charts, upcoming releases by country, and a dedicated set for Indian cinema.
- `search_imdb` with filtering, and `get_imdb_details` for one title by its IMDb id
- Credits — `get_directors`, `get_cast`, `get_writers`
- The charts — `get_top_250_movies`, `get_most_popular_movies`, `get_top_250_tv_shows`, `get_most_popular_tv_shows`, `get_top_box_office_us`
- Release calendars — `get_upcoming_releases` by country code, with `get_country_codes_for_upcoming_releases` to find the right code
- Indian cinema — `get_top_rated_malayalam_movies`, `get_trending_tamil_movies`, `get_trending_telugu_movies`, `get_top_rated_indian_movies`, `get_upcoming_indian_movies`
- Vocabulary lookups so your filters are valid — `get_types`, `get_genres`, `get_countries`, `get_languages`
A RapidAPI account subscribed to the IMDb API, with the key in `RAPID_API_KEY_IMDB` for stdio mode. Python 3.13 or higher; run `uv run imdb-server`, or use the Docker image, which serves HTTP on port 8081.
One command plus a key — npx -y @smithery/cli install @uzaysozen/imdb-mcp-server --client claude, then supply credentials