> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://developers.hive.co/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://developers.hive.co/_mcp/server.

# Rate Limits

## Limits

| Constraint             | Limit                              |
| ---------------------- | ---------------------------------- |
| Sustained request rate | 1 request/second (60 per minute)   |
| Burst request rate     | 2 requests/second (120 per minute) |
| Items per batch        | 50 max                             |
| Request body size      | 1 MB max                           |

## When you exceed a limit

| Scenario          | Response                   |
| ----------------- | -------------------------- |
| Too many requests | `429 Too Many Requests`    |
| Payload too large | `413 Payload Too Large`    |
| Invalid payload   | `422 Unprocessable Entity` |

## Batching strategy

* Send as few requests as possible, with as many records per request as the batch limit allows (up to 50). At 1 request per second sustained, one request of 50 records beats 50 requests of 1 record: batching is the difference between seconds and minutes for a large sync.
* Retry on network failures or `5xx` errors. On `429`, back off before retrying.
* Don't retry other `4xx` responses without fixing the request first.