Settings · pgjev

All settings are ordinary PostgreSQL GUCs: SET jev.<name> = ..., ALTER ROLE ... SET, ALTER DATABASE ... SET, or postgresql.conf.

Setting	Default	Meaning
`jev.api_key`	env `TYPESAFE_API_KEY`	TypeSafe API key
`jev.model`	`jev-latest`	Model name, or a pin such as `jev-1.13.0`
`jev.threshold`	`0.5`	Probability at which `jev()` returns true
`jev.batch_size`	`20`	Rows per API request. Accuracy drops measurably above ~20–25; see How it works
`jev.concurrency`	`16`	Parallel API requests; up to twice that many are queued ahead of the executor
`jev.max_prefetch_rows`	`5000`	How far past a cache miss the read-ahead scans to find the requested row, and how many skipped rows it keeps for later requests (memory bound)
`jev.notices`	`on`	Progress `NOTICE` per finished request and a summary per table: requests, tokens, estimated cost, time
`jev.api_url`	`https://api.typesafe.ai/v1/systemone`	Endpoint (proxies, mocks)
`jev.timeout`	`30`	Seconds per API request. Waits are interruptible: `statement_timeout` and cancel requests apply within 250 ms
`jev.keepalive`	`600`	Seconds a pooled API connection may sit idle before it is reconnected. The first request on a fresh connection is the slow one (TLS plus up to 1.5 s of server-side setup)
`jev.max_rows_per_statement`	`0` (off)	Abort if the statement would send more rows than this
`jev.max_chars_per_statement`	`0` (off)	Abort if the statement would send more characters of row data than this

sql

SET jev.api_key = 'your-key';
ALTER ROLE analyst SET jev.api_key = 'your-key';
SET jev.threshold = 0.7;
SET jev.max_rows_per_statement = 2000;

jev.api_key falls back to the TYPESAFE_API_KEY environment variable of the server process, not of your psql client.

The last two settings are spend guards for shared deployments. 0 means off. Large tables shows how to use them. How it works explains batch size, concurrency, and the cache.