Requests and responses

Site IDsCopy Anchor Link

Endpoints identify a site by its project ID, a UUID. To find your project IDs, call List projects:

[{ "id": "3235d25b-6008-452e-a453-62a5f2644ac1", "name": "Example Solar Farm" }]

Most endpoints take a single project_id. Get performance data takes one or more project_ids, which must all belong to the same organization.

Query parametersCopy Anchor Link

• To pass several values, repeat the parameter: ?project_ids=a&project_ids=b.
• Boolean parameters accept true, false, 1, or 0.

TimeCopy Anchor Link

Endpoints use two kinds of time, depending on what they return.

Performance and lossesCopy Anchor Link

Get performance data and List loss events work in the site's local time, so a day means the site's day.

• dt_start and dt_end are dates, such as 2026-05-25. The range includes both days in full.
• Timestamps in the response, such as ts, are the site's local time in the form YYYY-MM-DD HH:MM:SS, with no time zone.

Signals and alarmsCopy Anchor Link

The signals and alarms endpoints work in absolute time.

• Date-time parameters are ISO 8601 with a time zone offset, such as 2026-05-25T00:00:00Z or 2026-05-25T00:00:00-05:00.
• Timestamps in the response are UTC.

IntervalsCopy Anchor Link

Endpoints that aggregate over time take an interval: OneMinute, FiveMinutes, TenMinutes, FifteenMinutes, ThirtyMinutes, OneHour, OneDay, OneWeek, OneMonth, or OneYear. Get performance data also accepts Full, which returns one value for the whole range.

PaginationCopy Anchor Link

Endpoints that can return many rows use limit and offset:

1. Send your first request with offset=0, or leave it out.
2. If the response has exactly limit rows, more might exist. Increase offset by limit, and send the request again.
3. When a response has fewer than limit rows, you've reached the last page.

Each endpoint's reference page lists its default and maximum limit.

Read only what changedCopy Anchor Link

To keep a copy of alarms or signals up to date without reading everything again, use the endpoint's change parameters, such as updated_since for alarms or received_since for signals. Each reference page explains how to resume from the last row you read.

Response formatsCopy Anchor Link

Endpoints that return time series let you choose a format with the Accept header:

Arrow and Parquet are the most compact formats, and load directly into tools such as pandas, Polars, and DuckDB. Other endpoints return JSON.

ErrorsCopy Anchor Link

Errors return a 4xx or 5xx status code. Most error bodies are JSON with a message:

{
  "error": {
    "message": "The project doesn't exist, or your token can't access it."
  }
}

A request with invalid parameters returns 400 Bad Request, with the problems listed under error.issues:

{
  "error": {
    "target": "query",
    "issues": [{ "path": ["dt_start"], "message": "Invalid date" }]
  }
}

Rate limitsCopy Anchor Link

• Platform API (/platform/v1): 100 requests a minute for each user.
• Inbound sync (/integrations/v1): 300 requests a minute for each set of credentials.

When you exceed a limit, the API returns 429 Too Many Requests. To stay under it, request larger time ranges and pages rather than many small ones.