The lyre that plays fate.
A voice-driven AI Dungeon Master for tabletop role-playing. Speak, and the story answers back — the narrator in one voice, every NPC in their own, real dice on the table, and a party that can gather around a single laptop or join from anywhere in the world.
Already have an invite code? Sign in at go.wyrdlyre.app →
How it works
Open the link on your phone or laptop. Sign in. Speak. No install, no API keys to configure, no menus to fight.
Hold the mic button, or say "Ok DM…" for hands-free voice. Text fallback is always a keystroke away.
Claude reads the scene, remembers the campaign, honors your house rules, and writes the next turn in the mood you chose.
Narration in one voice, every NPC in their own — warm innkeeper, gruff watchman, cautious ranger — right there at the table with you.
Built for your table
The host opens a session on their laptop. Every other player taps an invite URL on their own phone and picks their character — the DM attributes their voice turns to that character automatically. Everyone sees the same chat, combat tracker, NPC log, and scene image in real time.
Or cluster around one device — Wyrdlyre's voice fingerprint enrollment lets the DM identify which human at the table just spoke, even when you're all sharing a mic.
When there's no one else around
Add AI companions to your campaign and the DM voices them for you — each with a persistent voice, a distinct personality, and their own reasons to speak up when the moment calls.
The stoic ranger you traveled with for three sessions will sound the same in session four. She'll argue with the weary monk. She'll warn you when she thinks you're being stupid. She might even be right.
Every campaign, its own voice
Pick the tone of your campaign once. The DM's narration, pacing, pauses, and the quality of the silence all adjust to match — and you can tune the DM's personality on a second axis, from Classic to Wry & Irreverent.
Plus Gritty Low-Fantasy, Saturday Morning Cartoon, and Noir Detective. One-click setting templates: Classic high fantasy, High-seas adventure, Apocalyptic wastes, and more.
Dice you can trust
Every roll in Wyrdlyre comes from a deterministic dice engine — audited per-die, operating-system entropy, every result inspectable. When the DM asks for a Dexterity save, the number on your screen is the real one. Not a number the AI decided would be dramatic.
And if your hand already reaches for the dice bag — roll your own. The dice panel takes any value you type. Real d20 on the table, real result in the app, real reaction from the DM. Whether your dice live on your screen or in your palm, the story listens the same way.
In the current alpha
Pricing, when we flip the switch
Alpha testers play free. These are the draft prices waiting behind the curtain — final tiers and limits set at launch.
$0 / month
A short campaign every month. Enough to see the magic.
$19.99 / mo
Regular weekly sessions for you and your party. The default.
$39.99 / mo
Unlimited play, ElevenLabs voice cloning, premium DM models.
$4.99 / mo
Bring your own Anthropic + OpenAI keys. We charge for the software, you pay providers direct.
Draft planned pricing. Final tiers may shift based on alpha feedback.
Where we are
The full voice loop ships: speak, the scene is read, the DM writes the next turn in character, narration and NPCs speak back in distinct voices, everyone at the party sees the same board state. Combat tracker, NPC log, scene images, character portraits, session transcripts, trash/undo — in testers' hands today.
Every tester session is instrumented: per-turn ratings, a session exit survey, and a bug button on every page. The feedback shapes what ships next.
If you'd like to join the next round of testers, the form below is how you raise your hand.
Questions we get asked
No. Wyrdlyre runs in your browser at go.wyrdlyre.app on any modern phone, tablet, or laptop. Sign in with Google, redeem your invite code, you're in. Chrome, Edge, and Safari work best; Firefox works with a slightly slower voice path.
Yes. Add AI companions during campaign setup — each gets a voice and a personality and becomes part of your party. They react to scenes, argue with each other, and sound consistent across sessions. Solo play is a first-class path, not an afterthought.
D&D 5e at launch, via the Creative Commons SRD 5.1. For 1st Edition AD&D, 2e, 3.5, and 4e — rulebooks still under copyright — the planned v2 "BYO rulebook" mode will let you upload your own PDFs for the DM to reference during play.
Alpha testers play free. When paid tiers turn on, four options — mapped to the plans above:
Both, equally. Multi-player works across devices (each player on their own phone) or around a single shared laptop (with voice-fingerprint speaker ID). Solo works with AI party companions. You can swap between modes in the same campaign — run one session with friends, the next on your own with the same characters.
Raw audio is transcribed and discarded — we don't retain the waveform. Transcripts of what was said are stored as part of the session log (so the DM has continuity) but you can export or delete any campaign at any time; deletion sends it to a 15-day trash you can restore from, then it's gone.
When the alpha cohort tells us it's ready. Human testers are running campaigns right now and every rating and piece of written feedback shapes the next build. If you want an early seat, the waitlist below is the path.
Waitlist & tester sign-up
Tell us where to email you when something's worth saying — and, if you'd like to test, a few questions that help us match you to the right tester cohort.
Your email client should have opened with your answers. Hit send and we'll take it from there — expect a reply within a day or two, or the next tester-seat batch, whichever comes first.