pvperf in AI assistants
An AI assistant runs pvperf through its MCP server or its REST API. Both take your pvperf API key.
Your API key
Sign in to the workbench with your email. Open the Account page and create a key. The key starts with pvp_live_. Each assistant sends it in the header Authorization: Bearer pvp_live_….
Runs are charged in units. Each month starts with free units. GET /v1/pricing gives the price per unit and the free units per month. A run that fails is not charged.
Setup pages
| Assistant | Connection | Page |
|---|---|---|
| Claude (claude.ai, Claude Desktop, Claude Code) | MCP server | pvperf in Claude |
| ChatGPT | GPT action | pvperf in ChatGPT |
| Cursor | MCP server | pvperf in Cursor |
| Your own code | REST API | Python example, TypeScript example |
The MCP server
The server's address is https://pvperf.com/mcp. Its transport is streamable HTTP. It takes the API key as a bearer key.
| Tool | What it does | Charged |
|---|---|---|
simulate_simple |
Runs a system from a location and a size. Returns the full request it ran. | Yes |
simulate |
Runs a full request, for example the edited request of a simple run. | Yes |
get_defaults |
Returns a complete request for a 5.4 MWp plant of fixed racks. | No |
get_typical_weather |
Returns the id and the source of the typical year at a location. | No |
get_run |
Reads one of your stored runs. | No |
list_runs |
Lists your stored runs, newest first. | No |
search_modules |
Searches the module library. | No |
search_inverters |
Searches the inverter library. | No |
add_component_file |
Stores a PAN, OND or weather file and returns its id. | No |
validate_scene |
Checks a 3D scene and counts its racks, objects and modules. | No |
describe_model |
Lists each step's method for a request on a path, and each method's variants. | No |
get_horizon |
Returns the terrain horizon at a location. | No |
get_docs |
Returns the documentation page of a field, step or guide. | No |
send_feedback |
Sends a bug report or feedback to the pvperf team. | No |
The unchanged request of get_defaults is the default project. Its run is free.
Other references
- llms.txt describes the API for language models.
- The API page describes the requests and the answers.
- The OpenAPI document gives every field with its unit.