# Paris Traffic and Taxi Stand AI Agent Connect > Paris Traffic: Road Counts & Taxi Stands provides direct access to the City of Paris open data. Your AI client can query hourly vehicle counts, mean speeds, and congestion states from 3,700 permanent counters, or locate any of the 420 official taxi stands across the city's arrondissements. ## Overview - **Category:** government-public-data - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_ddHxhojZtH6eau5ajOHUC5IEnAlojzEoQeRo7KxO/ai-agent-connect - **Tags:** paris, france, traffic, congestion, taxi, road-counters, keyless ## Description You can pull precise traffic metrics and transit locations directly into your AI agent. This MCP connects to the official Paris open data platform to give you granular visibility into the city's movement. Instead of digging through web portals, you can ask your agent to find specific road counters by name or code, check how many vehicles passed a certain intersection at a specific hour, or identify the nearest official taxi stand. The data includes vehicle counts, average speeds, and specific congestion states like "Saturé" or "Fluide." You can also query the 420 official taxi stands to find their address, bay count, and current status. Because the dataset is massive, the MCP is designed to handle specific site lookups and hourly windows, making it easy to analyze how long a specific corridor stays congested or where taxi capacity is highest in a given arrondissement. ## Tools ### count_taxi_stands Returns a total count of official taxi stands. You can filter this by arrondissement or current status to compare availability across different parts of the city. ### list_traffic_sites Provides a list of the ~3,700 permanent road-traffic counter sites in Paris. Each entry includes the site code, road name, and neighboring counter information. ### taxi_stand_lookup Finds specific taxi stands by their name. It returns the address, arrondissement, number of available bays, and the stand's current status. ### traffic_counts Retrieves hourly traffic measurements for a specific counter site. You get the vehicle count, mean speed, and the congestion state for a given time window. ### traffic_counts_by_state Counts how many hourly records fall into a specific congestion state. Use this to see how often a site hits "Saturé" or "Fluide" status. ### traffic_site_lookup Locates a specific traffic counter using its unique iu_ac code. It provides the road name and details about adjacent counters. ### list_taxi_stands Lists the official taxi stands across Paris. You can filter by arrondissement or status to get names, addresses, and bay counts. ## Prompt Examples **Prompt:** ``` Which traffic counter sits on Avenue de Gambetta, and how many cars passed at 8am today? ``` **Response:** ``` The counter for Av_Gambetta is located at site code [iu_ac]. At 08:00, the vehicle count was [q] with a mean speed of [k] km/h and a congestion state of [etat_trafic]. ``` **Prompt:** ``` How many hours did the counter at iu_ac 1232 stay saturated in the last week? ``` **Response:** ``` The counter at 1232 recorded [count] hours in the "Saturé" state during the requested period. ``` **Prompt:** ``` List the official taxi stands in the 8th arrondissement. ``` **Response:** ``` Here are the official taxi stands in the 8th arrondissement (75108): [List of names, addresses, and bay counts]. ``` ## Capabilities ### Traffic Monitoring Your agent uses this to pull hourly vehicle counts and speeds from specific road counters. ### Congestion Analysis The AI calculates how many hours a specific road stays in a saturated or fluid state. ### Taxi Stand Location Your agent finds the address and bay capacity of official taxi stands in any arrondissement. ### Site Identification The AI maps road names to specific permanent traffic counter codes. ## Use Cases ### Congestion Pattern Tracking An agent analyzes how many hours a specific corridor stays in a "Saturé" state during morning rush hour. ### Taxi Availability Checks A user asks for the number of available bays at taxi stands in a specific district. ### Route Impact Assessment A planner checks the mean speed at a specific road counter to evaluate recent traffic changes. ### Infrastructure Mapping An analyst uses site lookups to find neighboring traffic counters for a specific street. ## Benefits - Direct access to the Paris open data platform without manual searching. - Granular hourly traffic data for 3,700 specific locations. - Exact matching for congestion states and arrondissement codes. - Instant counts of taxi stand availability and capacity. ## How It Works Connect your client to Vinkius to access the Paris traffic and taxi data immediately. 1. Connect your MCP-compatible client to Vinkius. 2. Ask your agent a question about Paris traffic or taxi stands. 3. The agent selects the appropriate tool, such as traffic_counts or list_taxi_stands. 4. The MCP executes the request against the Paris open data platform. 5. Your agent receives the raw data and presents the answer to you. ## Frequently Asked Questions **How many traffic counters are available in this MCP?** The MCP provides access to approximately 3,700 permanent road-traffic counter sites across Paris. **Can I filter traffic data by specific congestion states?** Yes, you can filter by the exact states stored in the system: "Fluide", "Pré-saturé", "Saturé", or "Inconnu". **How do I find a taxi stand by its location?** You can use the list_taxi_stands tool to filter by arrondissement or use taxi_stand_lookup to find a stand by its name. **What kind of traffic metrics can my agent retrieve?** Your agent can retrieve hourly vehicle counts, mean speeds in km/h, and the current congestion state for any site. **Does this MCP require a separate API key for Paris Open Data?** No, Vinkius hosts the MCP and manages the connection to the Paris open data platform for you. **Do I need an API key?** No. opendata.paris.fr serves all city datasets anonymously over its Opendatasoft REST API. This MCP defines no credentials and needs nothing configured. **Why does traffic_counts require a site code?** The counter dataset holds ~29.6M hourly rows, and the ODS platform rejects a bare time-window query over it. Pass an iu_ac site code (from list_traffic_sites) and the window applies to that site. **What are the congestion states?** Four stored values: "Fluide" (free-flowing), "Pré-saturé" (near capacity), "Saturé" (saturated) and "Inconnu". Accents matter — use them exactly. **Are the taxi stands live?** No — the dataset is a static official reference of the ~420 stands (name, address, arrondissement, bay count, status), not a real-time occupancy feed.