It writes the SQL. You decide what runs.
Free · macOS 11+ · Apple Silicon & Intel
No pasting your schema into a chat window. pgen reads the live catalog itself and sends the right slice with every question.
Add a column this morning and it knows about it. Rename a table and the context follows.
There is no pgen server in the path.
Your schema and your question go straight from your machine to the provider you picked - we never see either. Point it at Ollama and nothing leaves the machine at all.
Anthropic, OpenAI, Google, a local Ollama, or any OpenAI-compatible endpoint you host yourself. Keys are stored in the operating system keychain, not in a config file, and never sent anywhere except the provider they belong to.
The safety layer was built before the assistant was. It applies to SQL the model wrote the same way it applies to SQL you wrote, and none of it can be disabled from a prompt.
SELECT c.id, c.name, sum(o.total_cents) …UPDATE invoices SET sent_at = now() …DROP TABLE customersALTER TABLE orders ADD COLUMN …The editor, query plans, connections, searchable history, import and export - the parts you'd use every day even if the AI didn't exist.
tabs and unsaved text survive a restart
Autocomplete drawn from the live catalog - your tables, your columns, your joins - not a keyword list. Browse table data in a grid that pages as you scroll, and edit cells in place - every change committed in a single transaction.
Every node with its cost and time, the expensive branch marked red, and a plain sentence telling you which index would fix it - instead of a plan you have to decode yourself.
Tag connections as Production, Staging, or Development and filter by environment. Your last 500 queries stay searchable - kept per connection, on your machine.
Column mapping with type inference and conflict resolution on insert. Export as CSV, JSON, SQL, or Excel - filtered by a WHERE clause when you don't need the whole table.
Today the assistant writes SQL and you decide what runs. Here's what's coming next - all of it arriving behind the same approvals you already control.
Bring up a table that didn't make it into the context and the assistant looks it up in the live catalog on its own - mid-conversation, before it answers.
The assistant can safely try a query before it gives it to you, so what lands in your editor is something it has already run - read-only, row-capped - and seen come back clean.
Point it at a query that's dragging and get a straight answer about what's holding it up - without having to read a query plan yourself.
Get a specific recommendation with the reasoning behind it, instead of guessing which column to index and hoping it helps.
If it speaks the wire protocol, pgen connects. Cloud, container, or laptop - paste the URL and you're in.
Free, every feature, no account. Add your own API key when you want the assistant on, or run it against a local model and skip the key entirely.
We use analytics to understand which pages get visited and how many people download pgen. See our privacy policy.
