A MongoDB IDE that
knows your schema

Autocomplete built from your own collections, that keeps following a field through $lookup, $group and $facet. A visual aggregation builder, live cluster metrics, and 70+ MCP tools for AI assistants.

Download Monghoul See what it does

platforms: Windows, macOS, and Linux
installer: 38 MB
version: v1.12.1
requires: MongoDB 4.4+
price: Free tier, no signup

A worked-in workspace: hovering a collection answers what is in it, opening it lands on a populated table, and a filter typed into the query already there is completed first with the field names the app learned from the documents and then with the values that field actually holds. A wrong total is corrected in the cell, a column sorted and another pinned while the table scrolls past it, and a date edited through a calendar in the tree. A pipeline typed in the code editor is offered a field that did not exist until the stage above invented it, runs, and is read as a chart. The chart tab is carried to the edge so the workspace splits, the query is saved, renamed and filed, and the whole thing is recoloured through three themes and one grown from scratch — The tour · What the editor knows, what the result lets you do, and whose workspace it is Pro

01 · Querying

Autocomplete that follows your fields through the stages that reshape them.

A Monaco editor over the MongoDB Node driver, with completions pulled from your actual collection schemas. The part other clients do not do is keep those completions correct deep inside a pipeline.

Querying → ▶ 0:45 demo

The autocomplete list open on a nested field path, each entry showing its type — Query editor · Typing "shipping." offers only the child fields, with types

02 · AI access

70+ MCP tools, with a permission model in front of them.

Monghoul runs an MCP server, so an AI assistant can query your databases through the same connection profiles you already use. The interesting design is not the tools. It is what an assistant is allowed to reach and what it has to ask for first.

AI access → ▶ 0:44 demo

The AI access panel showing the tool permission list and the approval setting — AI access · Scope and per-tool permissions, set before anything runs Pro

03 · Reading results

Six ways to read a result, and a performance grade on every read that returns a cursor.

Table, tree, JSON, result, explain and chart. The view you want depends on the question you are asking, so switching between them costs one click and no re-run.

Reading results → ▶ 0:51 demo

Query results in a virtualized table, sorted on durationMs, with the country column pinned to the left edge — Results · Table view, virtualized, with a pinned column and a sort

04 · Connections and safety

Eight authentication methods, TLS and SSH, and a block that lives at the driver.

This app holds credentials for your production databases. What it does with them, and what it stops you doing by accident, is worth reading before you install it.

Connections and safety → ▶ 0:39 demo

The connection dialog on its General tab, with a pasted connection string and the Security and Advanced tabs beside it — Connections · Paste a connection string, and the rest is behind two tabs

And the rest of it

3 more workflows, each with its own page and its own screenshots.

Free stays free

It is not a trial. The free tier is a usable MongoDB IDE with 8 connections and no time limit. Pro removes the caps and adds charts, monitoring and AI access, from $9 /mo, or once as a lifetime licence.

Compare Free and Pro →

Saved connections: Free 8

Pro Unlimited
Authentication: Free Password and SCRAM

Pro Plus X.509, LDAP, Kerberos, AWS IAM
Schema sampling: Free 500 documents

Pro Unlimited
Charts, monitoring, AI access: Free Not included

Pro Included

Capability	Free	Pro
Saved connections	8	Unlimited
Authentication	Password and SCRAM	Plus X.509, LDAP, Kerberos, AWS IAM
Schema sampling	500 documents	Unlimited
Charts, monitoring, AI access	Not included	Included

Point it at a database

38 MB installer · Windows, macOS, Linux · 14-day Pro trial, no card

Download Monghoul