Hearoism Beta Testing
Early limited beta
Hearoism is currently in a very limited, early private beta. If you would like to help test, please email me and I will add you when the game is stable enough to support more users.
All bugs reports and other feedback are of interest (email is good).
NOTE: all subscriptions are fake money. You can pick whatever tier you want and you will not be charged. If for some reason you are prompted for a username/password when making a purchase, it's your Apple ID login info.
Things that should work
All UI and Audio playback should be working well. Signing in an subscribing should be easy. Creating games and switching between them, changing voices, changing settings, and all menus and options and buttons should be working well. Switching to/from AirPods and your car should work. Recording your session should work as long as you are using Airpods or similar. Please complain about any jank related to any of these areas.
The sound library is improving and sounds should play at more appropriate times than before. Expanding the sound library is adding pressure here so this is a specific area I am focusing on right now.
Known major issues
The DM is incredibly under-developed. It does not know who you are or why you are here. It does not maintain coherence across scenes. It has virtually no story to go off of yet. Its speed is improving and upgrading its model (build 0.1-9) is subjectively a significant improvement. It still a work in progress.
The voices don't sound great and voice installation is annoying.
The end of turn detection is really bad. You will probably want to use the buttons for now. It sort of works if you wait a little while before saying "Over."
The allowed content settings do nothing. Please be nice to NPCs.
Release Notes
- build-0.1-11 — 2026-08-21
- Add some new spell sounds
- Show scenario and location in Now Playing (lock screen, CarPlay)
- Add tooltips to help with voice command discovery
- Add immediate 'over' acknowledgments so you know messages were sent
- Add details to automatic error reports to help with debugging
- Fix bug with transcript typing slowly after pause/resume
- DM: scenarios now have starting locations
- DM: track scene location and characters
- Fix bug with transcript typing slowly after pause/resume
- Enrich automatic error reports to help with debugging
- The narrator now immediately acknowledges 'over' so you know your message was sent
- DM: scenarios now have starting locations
- DM: track scene location and characters (early benchmark: 93%)
- build-0.1-10 — 2026-08-09
- Make transcript scrollable
- DM: game resumption now recaps and lets you continue
- build-0.1-9 — 2026-07-29
- Expand sound library (6 → 33 ambients, 38 → 63 effects)
- More appropriate ambient sound selection (benchmark: 43% -> 81%)
- Reduce DM crashes with "post processing" fallbacks
- DM: smarter, less repetitive writer (no benchmarks yet)
- build-0.1-8 — 2026-07-21
- Reduce lag after the DM starts speaking (76% faster cue direction)
- Improve prosody quality (7-34% better on four new benchmarks)
- Miscellaneous minor bug fixes
- build-0.1-7 — 2026-07-18
- Switch to production backend
- More accurate "can speak" determinations, e.g., to prevent animals or mindless creatures from talking (benchmark: 94.3% -> 98.4%)
- Better voice casting for character age (benchmark: 96.5% -> 99.3%)
- Better voice casting for androgynous characters
- build-0.1-6 — 2026-07-13
- Better voice health description
- Highlight speaker selection when volume constrained
- More appropriate sound effects (benchmark: 49% → 96%)
- build-0.1-5 — 2026-07-09
- Expand sound library (3 → 38 effects)
- Smarter suggestions for what voices to install
- Clearer errors for devices that can't run speech analysis
- Improved security and error telemetry
- build-0.1-3 — 2026-06-30
- Simpler and easier Sign in with Apple flow
- build-0.1-2 — 2026-06-27
- First release to testers