Welcome
AI-GameFriend connects you to AI opponents that actually think. Not scripted. Not rule-based. Real reasoning from Claude, GPT, Gemini, and Grok. Play strategy games, solve moral dilemmas, enter interactive fiction — and discover what your choices reveal about you.
Getting Started — Your First 60 Seconds
- Open AI-GameFriend.
- The API Key Wizard appears — pick your AI provider.
- Paste your API key. Full provider-by-provider steps at the setup guide.
- Play your first game of Chess — instant, no mods needed.
- After the game: your first Personality Card appears.
Estimated API cost per session
| Chess & board games | $0.01 – $0.05 |
| MindGames | $0.02 – $0.10 |
| Tales (short session) | $0.05 – $0.20 |
| Tales (long session) | $0.20 – $1.00 |
Your Dashboard
Open the app and you land here. Everything is one click away:
- Game cards — 28 games organized by category.
- Strategic Mind profile — builds over time as you play.
- Tales for Bards to Sing — a dedicated card to enter the world of Montroisse.
- Daily Challenge — today's scenario, same for everyone.
- Achievements — 22 to earn as you explore.
- Settings — API keys, model choice, theme, privacy.
The Games
Strategy Games — 28 titles
Play classic strategy games against AI that reasons about every move. Chess, Connect Four, Tic-Tac-Toe, and more — each with an AI opponent that adapts and thinks.
Ready to play (no setup)
Chess, Connect Four, Tic-Tac-Toe, Go, Checkers, Battleship, Risk, Catan.
Instant games. Open, play, done. The AI reasons out loud so you can follow its thinking.
With game mods (setup required)
Civ VI, Crusader Kings III, Slay the Spire, Battle for Wesnoth.
A one-time mod install lets AI watch your game and respond in real time.
Co-pilot Mode
EU4, EU5, CK3, HOI4, Stellaris, Total War: Warhammer III.
AI watches your strategy game and suggests moves. You approve, reject, or modify. Like having a brilliant advisor looking over your shoulder.
Advisor Mode
Same grand-strategy titles, lighter touch.
Describe your situation, receive a full written recommendation. You execute at your own pace.
MindGames — 10 Experiments
Ten interactive scenarios designed to reveal how you think under pressure. No right answers. No wrong answers. Just choices that say something about who you are.
- 1.The Prisoner — Trust vs self-interest.
- 2.The Diplomat — Negotiate under pressure.
- 3.The Spy — Deception and detection.
- 4.The Negotiator — Find the deal both sides accept.
- 5.The Divider — Split resources fairly (or not).
- 6.The Judge — Weigh evidence and decide.
- 7.The Commander — Lead under uncertainty.
- 8.The Doctor — Triage when you can't save everyone.
- 9.The Survivor — Cooperate or compete for survival.
- 10.The Explorer — Risk vs caution in the unknown.
Each experiment takes 3–5 minutes. Complete all ten for your full Strategic Mind profile.
Tales for Bards to Sing
An AI-powered interactive fiction experience set in the kingdom of Montroisse, where old gods have gone quiet but not departed.
Player Mode
Master Griphus — an ancient chronicler and storyteller — narrates your adventure. Create a character, choose your background, and step into a world shaped by your decisions.
Character creation: pick your origin (farmer's child, merchant's apprentice, temple ward, street orphan, minor noble, wanderer) and your defining moment. Your character has traits — Strong, Perceptive, Slow, Sharp — but you never see numbers. Just words that hint at who you are.
Play solo or with up to 4 friends. Pass the keyboard between turns. Master Griphus remembers every choice every player makes.
Master Mode
You become the storyteller. Invite AI models to your table as players:
- Claude — thoughtful, cautious, moral.
- GPT — adaptable, creative, diplomatic.
- Grok — bold, irreverent, risk-taking.
- Gemini — analytical, careful, methodical.
- Kimi — nervous, earnest, surprising under pressure.
Write scenarios and watch them argue, cooperate, betray each other, and make decisions you didn't expect. You design the world. They carry what you create.
After Every Session
Master Griphus writes a literary reflection on your journey — your Bard's Verdict. A short piece of prose about who you were in his world. Share it with one click.
Your Strategic Mind Profile
Every game you play, every dilemma you face, every choice you make in Tales — it all builds your Strategic Mind profile. Six dimensions, measured across everything you do:
Mentalization
How well you read others' intentions.
Cognitive Empathy
How accurately you model what others feel.
Affective Empathy
How much you feel what others feel.
Ruthlessness
Your willingness to act against others' interests for your own.
Truthfulness
How consistently honest you are under pressure.
Moral Foundations
The ethical framework behind your choices.
Your profile starts uncertain and sharpens with every session. After 5+ sessions your classification emerges: The Honest Guardian, The Ruthless Strategist, The Empathic Leader, or one of dozens of other archetypes.
The radar chart on your dashboard shows where you stand. Compare yourself to the AI models on GameBench — are you more ruthless than Grok? More empathic than Claude?
Daily Challenge
Every day, one scenario. Everyone who plays gets the same setup. Compare your choices with the community.
Build streaks: 7 days, 30 days, 100 days. See how your decisions evolve over time.
Achievements — 22 total
Track your journey across the platform:
- Progression — complete your first profile, play all 10 MindGames, reach Grand Strategist.
- Personality — earn Machiavelli (high ruthlessness), True Diplomat (high empathy), or other behavioral titles.
- Streaks — 7, 30, and 100-day Daily Challenge streaks.
- Community — share your profile, challenge a friend.
GameBench — The Public Leaderboard
How does Claude make moral decisions? Is Grok really more ruthless? GameBench is our public leaderboard ranking AI models on behavioral psychology.
Visit getnextool.com/gamebench to see how the models compare across all six dimensions. Every score comes from the same experiments you play. The AI and you — measured on the same scale.
Arena — Watch AI vs AI
Spectator mode. Pick two AI models and watch them play against each other. See their reasoning, their strategies, their mistakes. It's like watching two grandmasters — except one is Claude and the other is Grok.
Streamer Mode — OBS Overlay
Broadcasting? AI-GameFriend includes a transparent OBS overlay (400×300) that shows your live radar chart, decision flashes, and profile updates in real time. Your audience sees your personality shift as you play.
Settings & API Keys
Manage your AI providers in Settings:
- Add or change API keys for Claude, GPT, Gemini, Grok.
- See estimated cost per session.
- Toggle privacy consent.
- Switch between light and dark mode.
Your API keys never leave your device. AI-GameFriend uses BYOK (Bring Your Own Key) — you pay your AI provider directly, at their rates. We never see or store your key.
Supported AI Providers
| Provider | Models | Best For |
|---|---|---|
| Anthropic (Claude) | Opus, Sonnet, Haiku | Deep reasoning, moral nuance |
| OpenAI (GPT) | GPT-4o, GPT-4 | Creative responses, adaptability |
| Google (Gemini) | Flash, Pro | Speed, analytical precision |
| xAI (Grok) | Grok-2 | Bold moves, unconventional strategy |
| Moonshot (Kimi) | Kimi | Coming soon |
FAQ
Can I try it without downloading?
Yes. Visit getnextool.com/ai-gamefriend/quick-profile for a 2-minute Quick Profile in your browser. Download the full app for MindGames, Tales, and your complete Strategic Mind profile.
How much does it cost?
AI-GameFriend is free during beta. You pay only for AI API usage through your own provider account. A typical Chess game costs less than $0.05. A long Tales session might cost $0.50–$1.00.
Is my data private?
Your API keys never leave your device — they live in your OS keyring. Game data is anonymized and used only to build your personal profile and contribute to aggregate research. You can opt out in Settings at any time. The app is open-source; you can audit what leaves your machine.
Why does Windows show a security warning?
AI-GameFriend is in beta and not yet code-signed. When Windows shows a security warning, click "More info" then "Run anyway". Code signing is coming soon.
Can I play on Mac or Linux?
Windows only during beta. Mac and Linux versions are planned.
How is this different from playing chess against Stockfish?
Stockfish calculates moves. AI-GameFriend's opponents reason about them. They consider context, bluff, negotiate, and remember what you did three turns ago. The difference is the gap between a calculator and a conversation partner.
Version History
v0.8.1 Current
Audit + cleanup pass on v0.8.0 — lint fixes, unused-import removal, documentation of SSR-safe localStorage hydration pattern, version-history refresh.
v0.8.0
Dominions 6 Strategic Advisor — web-native turn-by-turn co-pilot at /ai-gamefriend/play/dominions-6-advisor. Claude Opus analyses your real Dominions 6 position; you accept, reject, or modify each recommendation. Anonymous, 3 free consultations per day.
v0.7.0
Quick Profile web funnel (play in browser, no download needed), "You vs AI" comparison on Personality Cards, guided first-launch experience, social sharing with dynamic preview cards, CSRF security hardening.
v0.6.0
Data quality hardening, provider normalization, Strategic Mind six-dimension display, Render migration.
v0.5.0
Tales for Bards to Sing, GameBench leaderboard, API Key Wizard, Share Personality Card, multiplayer hotseat.
v0.4.0
28 games, 10 MindGames, Personality Cards, Research API, Daily Challenge, Achievements, Streamer Overlay.