# Solar Energy PVGIS AI Agent Connect > Solar Energy connects your AI client to the European Commission's PVGIS database. It provides monthly and yearly electricity yield estimates, optimal tilt and orientation calculations, and detailed meteorological profiles for any location on Earth. Use it to size batteries, compare mounting configurations, and understand the specific solar potential of a site without needing API keys. ## Overview - **Category:** energy-utilities - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_WzfWNt705fVjTNHaEH9UkmI37PsdwZu97F3qkUwX/ai-agent-connect - **Tags:** solar, photovoltaic, renewable-energy, energy, pv, pvgis, solar-panels, rooftop, irradiance, climate ## Description You can now give your AI agent direct access to the Photovoltaic Geographical Information System (PVGIS) maintained by the European Commission. This MCP pulls from satellite and reanalysis weather records to give you high-fidelity solar data for any coordinate on the planet. Instead of manually toggling through web interfaces, you can ask your agent to run complex comparisons or calculate specific energy outputs for a proposed installation. When you ask for a yield estimate, the MCP doesn't just give you a single number. It provides the tilt, orientation, and system losses used to reach that figure, so you know exactly what assumptions are baked into the result. You can use it to decide if a fixed roof mount is more cost-effective than a purpose-built frame by comparing multiple configurations in a single request. For battery sizing, you can pull hourly power profiles to see exactly when energy arrives during a specific month. It handles the heavy lifting of processing 8,760 hourly records into usable monthly averages, making it a practical tool for real-world solar design and site assessment. ## Tools ### estimate_pv_generation This tool calculates monthly and yearly kilowatt-hours for any site using specific latitude and longitude. It scales linearly with the peak power you provide and can automatically select the best tilt and orientation for the location. ### find_optimal_tilt This tool identifies the best slope and azimuth for solar panels at a specific coordinate. It returns the resulting yearly and monthly energy yield for that ideal fixed mounting. ### compare_configurations This tool compares multiple mounting setups, such as a flat roof versus a tilted frame, in a single call. It sorts the results by yearly yield so you can see exactly how much energy you lose by choosing a cheaper mounting option. ### get_monthly_profile This tool provides an hourly breakdown of power delivery for a specific month. It aggregates the annual hourly series into a mean daily profile to help you understand when energy is actually available. ### get_typical_meteorological_year This tool retrieves monthly temperature, humidity, wind, and pressure data along with sunshine splits. It uses long-term climate records to provide a representative meteorological year for a location. ## Prompt Examples **Prompt:** ``` My roof in Lisbon has a 25-degree pitch facing southeast. What would a 6 kWp solar system generate there, and is it worth re-tilting the panels? ``` **Response:** ``` A 6 kWp system on a 25-degree southeast-facing roof in Lisbon would generate about 9,650 kWh a year. PVGIS suggests an optimal tilt of 35 degrees facing south would yield about 10,060 kWh. Since the roof is already within 4% of the maximum, re-tilting would only gain you about 400 kWh per year, which likely won't cover the cost of new mounting hardware. ``` **Prompt:** ``` I'm sizing a battery for a cabin near Milan. When in the day does a 5 kWp array actually deliver power in January? ``` **Response:** ``` In January near Milan, a 5 kWp array on an optimal tilt produces about 350 kWh for the month. The hourly profile shows power arriving between 8am and 5pm, with a midday peak near 1,900 W. You'll need to size your battery to handle the midday surplus and ensure it can cover evening loads when the output drops to zero. ``` **Prompt:** ``` Compare a flat mounting against a 30-degree south-facing frame for a site in Cape Town. ``` **Response:** ``` In Cape Town, a flat mount will yield less than a 30-degree south-facing frame (azimuth 180). The tool compares these configurations and sorts them by yearly kWh, allowing you to see the exact energy penalty of staying flat versus installing a frame. ``` ## Capabilities ### Yield Estimation The AI calculates monthly and yearly kilowatt-hours for any system size at any global coordinate. ### Mounting Optimization Your agent finds the ideal tilt and azimuth to maximize energy production for a specific latitude. ### Configuration Comparison The AI weighs different mounting options side by side to find the best balance of yield and installation cost. ### Hourly Power Profiling Your agent analyzes hourly output to help you plan battery storage and self-consumption strategies. ### Climate Analysis The AI retrieves typical meteorological data including temperature, humidity, and sunshine types. ## Use Cases ### Roof Suitability Checks Compare a standard roof pitch against an optimized tilt to see if the energy gain justifies the hardware cost. ### Battery Sizing Analyze the hourly power delivery in winter months to ensure storage capacity meets evening demand. ### Site Selection Use meteorological data to understand the sunshine split between direct and diffuse light at a new location. ### System Scaling Quickly scale energy estimates from a 1 kWp baseline to much larger commercial arrays. ## Benefits - Accesses official European Commission PVGIS data without needing to manage API keys. - Provides the specific assumptions like tilt and losses used for every calculation. - Enables direct comparison of multiple mounting angles in a single request. - Uses long-term weather averages from 2005 to 2023 for reliable modeling. ## How It Works Connect your AI client to the Vinkius-hosted MCP and start querying solar data immediately. 1. Connect your MCP-compatible client like Claude or Cursor to Vinkius. 2. Provide the latitude and longitude of your target site to your AI agent. 3. Ask the agent to estimate yield, find optimal tilt, or compare configurations. 4. The agent calls the PVGIS tools and returns the calculated energy data and climate profiles. ## Frequently Asked Questions **Do I need an API key?** No. PVGIS is a free public service of the European Commission Joint Research Centre and answers plain requests with no key, account or registration. **Which locations are covered?** Everywhere on Earth. The default database, PVGIS-ERA5, is global. PVGIS-SARAH3 is offered as an alternative and is built from satellite observations over Europe, Africa and Asia; requesting it for the Americas is rejected by PVGIS itself, which names the valid values in the error message. **Are these predictions for the coming year?** No, and this distinction matters. PVGIS reports the long-term average from its weather record — typically 2005 to 2023 — so the yearly figure is what a system of that size at that orientation should produce in an average year, and the variability field shows how much real years scatter around it. Next year will be sunnier or cloudier than that; the typical-meteorological-year tool exists for exactly this reason, to describe the climate rather than forecast a season. **Can it model solar tracking systems or concentrate the analysis on one module technology?** Not in this API. The tools model a fixed free-standing array with a stated tilt and orientation; PVGIS accepts a tracking parameter in this version but ignores it, so no tracking option is exposed rather than return a silently wrong figure. Module technology is likewise treated as crystalline silicon, which is what the standard loss assumptions assume. **Why does the monthly profile say a month is only partly clear when the sun shone all week?** The clear-sky share is derived from the irradiance the site itself received, hour by hour, compared against the clearest that hour reached anywhere in the month — PVGIS's own clear-sky flag exists in the raw data but reads zero for every hour in this version, so it cannot be trusted. The threshold counts an hour as near-clear at 70% of that hour's monthly best. It describes how sunny the month was relative to its own ceiling, not an absolute sky condition, and it is a supporting detail next to the kilowatt-hour totals rather than a measurement in its own right.