See every relationship in one graph
The web of relationships among characters, places, organizations, and items, auto-arranged by a force-directed graph — even the most tangled relationships become clear at a glance.
Get started nowLitMemo's relationship graph uses a force-directed layout to auto-arrange the network of relationships among characters, places, organizations, and items — closely tied nodes naturally cluster, so core groups and peripheral characters are told apart at a glance. As the cast grows and master-apprentice, lovers, rivals, and allies interweave, your head simply can't hold it all, and one slip writes a contradiction like 'friends earlier, enemies later.' The graph lays every relationship out on one image, and lets you filter by entity and relationship type and search for nodes. That's the biggest difference from hand-drawn relationship charts and sheer memory.
"Wait, what exactly is the relationship between these two characters? They seemed like friends earlier, then became enemies later..."
The more characters and the more complex the relationships, the less your head can hold. Master-apprentice, lovers, rivals, allies — with relationships planted everywhere, one slip writes a contradiction.
The core features of Character relationship graph — visualize your character and worldbuilding network
Force-directed auto-layout
The graph uses the d3-force force-directed algorithm to automatically compute each node's optimal position, so you never drag lines and arrange by hand — closely related characters are 'pulled' together, and distant ones naturally drift apart. This delivers a free insight: where nodes cluster is the story's core inner circle, and those scattered on the fringe are the peripheral cast, so an ensemble's power structure and factional layout surface on their own without deliberate analysis. Different entity types are marked in different colors too, so characters, places, and organizations are told apart on the graph at a glance — and unlike a hand-drawn chart that needs redrawing entirely with one change, adding a relationship here re-arranges the layout automatically.
- Auto-layout, no manual arranging
- Relationship density sets node distance
- Node colors distinguish entity types
Multi-dimensional filtering and search
As the cast grows, a relationship graph packed with nodes and lines gets hard to focus. Multi-dimensional filtering lets you see only the side you want: filter by entity type (say, show only characters and organizations, hiding places and items) or by relationship type (say, only master-apprentice lines, only hostile relationships). To find a specific character, node search with autocomplete highlights and locates it on the graph after a few keystrokes. This means the same base graph can switch into countless perspectives to suit your needs — one filter to untangle romance lines, another to analyze the power map — so even a vast network can focus down to the very thread you care about right now.
- Filter the display by entity type
- Filter by relationship type
- Node search with autocomplete
World-wide global graph
The world-level relationship graph links every book inside it, presenting the whole universe's relationship picture rather than just one book's network. A built-in volume switcher lets you move freely between 'whole world' and 'single book' — see the full graph for how cross-book characters connect, or switch to one volume to concentrate on its relationships. The same character's status differences across volumes are reflected on the graph too. For authors of multi-volume runs, this is the only vantage point for surveying the whole story universe's character network — relationship lines planted across books, grudges and bonds spanning multiple volumes, all seen clearly in a single graph.
- Cross-book character relationships all in view
- Volume switcher for precise filtering
- Character status differences across books at a glance
The addressing matrix
The graph answers who is connected to whom. The matrix answers a different question: what they call each other. A character does not have one way of speaking — they have a different word for every person they speak to: deferential to a mentor, blunt with an old friend, a full name for an enemy. That does not fit in a field on a character sheet, because it does not belong to a character at all; it belongs to a pair. The matrix lays the cast out on an N×N grid where each cell holds the word belonging to that end, and the diagonal holds what a character calls themselves. It is the same data as the graph, seen differently: fill in a cell and a line appears on the graph.
- A character-by-character grid, up to 60 per axis
- Each cell is that end's own word; the diagonal is self-reference
- Same data as the graph, which can label each end's word
Have you run into this too?
「Two characters are friends earlier and enemies later, and you've muddled it yourself」
Build the relationships between characters as lines on the graph, so who relates to whom and how is all laid out on one image, and a glance before writing an interaction makes it clear. However tangled the web of relationships, you won't write it into self-contradiction, and you won't have two who should have fallen out chumming it up again.
「An ensemble cast grows, and you can't tell who's core and who's a peripheral side character」
The force-directed layout naturally clusters densely connected characters into a group and scatters loosely connected ones to the fringe, so the difference between the core group and the peripheral cast is clear at a glance. Which character is actually an invisible hub, which side character's screen time could be trimmed — the answer is in how densely the nodes cluster.
「The character relationships across three books in one world interweave, and you want to see the whole universe」
The world-wide global graph strings every book's character relationships into one big web, with a built-in volume switcher — survey the whole world's character connections, or focus on one volume, however you like. The grudges and bonds spanning multiple volumes, and the relationship lines planted across books, are all taken in within a single graph.
The usual way vs LitMemo
| The usual way | LitMemo | |
|---|---|---|
| How relationships are shown | Hand-drawn or memorized | Force-directed graph, auto-arranged |
| Core characters | Judged by gut feel | Node density tells at a glance |
| Focusing on specific relationships | Squint at the whole graph at once | Filter by entity / relationship type |
| Locating a node | Hunt through the tangle of lines | Node search with autocomplete |
| The cross-book picture | View each book separately | World-wide global graph links it all |
Get started in four steps
- 1
Create characters and relationships
Have character cards first and draw lines to build their relationships
- 2
Open the relationship graph
Enter the graph page, and the force-directed layout arranges the nodes automatically
- 3
Filter to focus
Filter by entity or relationship type, and search to locate quickly
- 4
Switch the world view
Use the volume switcher to move between a single volume and the whole world
Frequently asked questions
It supports all entity types — characters, places, organizations, items, races, rules, timeline events, scenes, chapters, and more. Each type is marked in a different color, so characters, places, and organizations are told apart on the graph at a glance. This means the graph shows not just interpersonal relationships but the whole world — who belongs to which organization, which character is tied to which place, all strung into the same web.
Yes. The graph offers multi-dimensional filtering — filter by entity type (say, show only characters and organizations, hiding places and items), or filter by relationship type (say, only master-apprentice lines, only hostile relationships). The same base graph can switch into countless perspectives: one filter to untangle romance lines, another to analyze the power map — and paired with node search with autocomplete, even a vast network can focus down to the thread you care about right now.
A single book's graph shows that book's relationship network, while the world-level graph links every book inside the world, presenting the whole universe's relationship picture. A built-in volume switcher lets you move freely between 'whole world' and 'single book,' the same character's status differences across volumes are reflected on the graph, and grudges and bonds spanning multiple volumes are all seen clearly in a single graph.
No. The force-directed algorithm computes the optimal layout automatically, naturally clustering closely related nodes into a group and scattering distant ones to the fringe, so the difference between the core group and the peripheral cast is clear at a glance. Which character is actually an invisible hub — the answer is in how densely the nodes cluster — and paired with type filtering and search, relationships untangle fast no matter how large the cast.
It is the same data, viewed differently. The graph shows structure — who connects to whom, and how the cast clusters. The matrix shows the words — what each end calls the other. Filling in a cell creates the relation, so a line appears on the graph immediately, and a relation built on the graph shows up in the matching cell. Turn on "Show each end's word" on the graph and each word is drawn beside the node it belongs to. The matrix is characters only for now — locations and organisations still appear on the graph, they just do not take a cell in the matrix.
Better together
Character management
Characters are the graph's core nodes — structured management is a must
Learn moreWorldbuilding management
Places, organizations, items — the other half of the graph's nodes
Learn moreCross-book sharing
A cross-book relationship graph linking the whole universe's character network
Learn moreWherever the pen rests, there is home
Your world
deserves somewhere to live
Start now, and scale up as your world grows