AI diagram generator

AI flowchart generator.

For a process with decisions in it, where the question is what happens next rather than what talks to what.

Download for macOS
v0.1.33 · .dmg · Apple Silicon & Intel
Signed & notarized by Apple · opens without a Gatekeeper warning
sha256 698955a0187bc039f4c74f5d05a9f10fbb27376a45788a0a241d1326b73873c7
Download for Windows instead
$curl -fsSL https://lucidtrain.com/install.sh | sh

What a flowchart shows

One process, as a sequence of steps and decisions, from a start to one or more ends. It is the right diagram for onboarding flows, approval chains, incident runbooks and any branching logic that people need to follow. It is the wrong diagram for anything with several independent actors, where a sequence diagram keeps the participants distinct and a flowchart flattens them into one line of control.

How Lucid Train draws one

Describe what you want, or point it at a repository. The agent emits a semantic graph of nodes and edges with no coordinates, and the ELK layout engine computes positions and orthogonal routing. That separation is the reason the output cannot come back with boxes overlapping or arrows crossing through cards, which is the characteristic failure when a language model is asked to place things on a canvas itself. It also means editing is editing the graph: change a component and the layout re-solves rather than drifting out of alignment.

Left to right, because branches need horizontal room

Flowcharts are laid out horizontally rather than in tiers, since a decision produces parallel branches that need to sit beside each other. The layout engine handles the routing, which is the part that goes wrong when a flowchart is drawn by hand: branches that rejoin several steps later produce long edges that a person routes badly and a solver routes well.

Every decision needs both branches labelled

The most common defect in a hand-drawn flowchart is an unlabelled or missing branch, which leaves the reader unable to follow one of the paths. Labelling both outcomes of every decision is a small discipline that makes the difference between a diagram someone can act on and a picture of a process.

Local, offline, your models

It runs as a desktop application. Against a local Ollama model the whole thing works offline with no API key and no telemetry, which is the practical difference from every browser-based diagram tool: your architecture never leaves the machine.

Flowchart notation

ShapeMeaning
Rounded rectangleStart or end
RectangleA step or action
DiamondA decision, with labelled outgoing branches
ParallelogramInput or output
CylinderA data store
CircleA connector to elsewhere in the chart
ArrowFlow of control from one step to the next

Try it

shell
$ Draw a flowchart for our incident response process: detection, triage, severity decision, escalation paths and resolution.
FAQ

Questions

Flowchart or sequence diagram?

Flowchart for one process with decisions; sequence diagram for several participants interacting over time. If you find yourself adding swimlanes to a flowchart to keep actors apart, you probably wanted a sequence diagram.

Can it do swimlanes?

Yes, as sections, which groups steps by whoever is responsible. That is often the version people want for an approval process where the handoffs are the point.

Can it turn existing code into a flowchart?

For a self-contained function or a state machine, yes and usefully. For a whole application, no: control flow across an entire codebase is not a diagram anyone can read.

Does it export to draw.io?

It exports SVG, PNG and Excalidraw scenes. Excalidraw is the editable format here rather than draw.io, so a team standardised on draw.io would be importing an image.

Related