Why is local AI better than using a cloud API?

Local AI advantages: Zero cost per player - no API or infrastructure bills. Offline capable - your game keeps working without an internet connection. Private by design - no personal data leaves the device. Cloud-based AI adds legal and UX risks: Immersion-breaking outages during play sessions. Legal compliance challenges (GDPR, AI Act, export laws). Tryll's local-first approach is not just cheaper - it's safer, faster, and scalable to millions of players.

How does Tryll optimize for different hardware configurations?

Our LLM Manager automatically: Detects hardware on first launch. Picks the most efficient model (4B, 8B, 14B). Adjusts memory usage and context size dynamically. Offloads high-level reasoning and memory management to modular agents (Director, Writer). Updates models automatically as the ecosystem evolves. This ensures smooth AI performance across a wide range of gaming machines.

What's the install size and memory footprint?

Default models range from 2-8GB download size. In memory, they typically use 4-12GB of VRAM depending on model size and quantization. The engine itself adds minimal overhead (~100MB). Players can choose smaller models for lower-end systems or larger ones for more sophisticated AI behavior.

Tryll Engine - On-Device AI for Games

HOW IT WORKS

Start Simple. Go Deeper When Ready.

Three steps to an intelligent assistant for your game.

Tell Us Your Game

Share your game name and a link to your wiki, docs, or knowledge base. Or we'll help gather it.

We Do the Heavy Lifting

No coding needed. We compile your data with source references, create and test prompts, validate for accuracy. You review, we iterate.

Players Stay Immersed

No more alt-tabbing to Google. No Reddit searches. No wiki rabbit holes. Answers appear instantly inside your game.

Ready for more? Upgrade to Companion for real-time awareness, or Engine Integration for custom in-game characters.

See Pricing

WHY TRYLL

We Handle the Hard Parts

You focus on your game. We take care of everything else.

Zero Infrastructure

No cloud APIs, no server costs, no rate limits. Tryll runs entirely on player hardware - you ship your game, not a data center.

100% local. Zero ongoing costs.

Prompts Done For You

We craft, test, and refine prompts that work. Built-in guard rails, hallucination detection, and prompt injection protection. You review, we iterate.

Battle-tested prompts. Your approval.

Start Without Code

Start with our overlay - no code changes needed. When you're ready for deeper features, add engine integration at your own pace.

Zero to launch. Grow when ready.

See Tryll Engine in Action

Discover how Tryll transforms your game with intelligent AI features

Natural Language Tutorials

Reduce support tickets and negative reviews. Connect your documentation, wikis, and guides to an AI that answers player questions instantly - no more frustrated players abandoning your game because they can't figure something out.

Relationship-Building AI

Create companions with persistent memory and evolving personalities. Your NPCs can remember past interactions, form opinions, and develop relationships - without writing thousands of dialogue branches.

AI That Takes Action

Give AI direct control over game mechanics. Let NPCs spawn enemies, modify environments, or distribute rewards based on context. Ship dynamic gameplay without hardcoding every possibility.

Dynamic World Generation

Generate quests, lore, and world events on the fly. Your game world can create infinite content based on player actions and preferences - no more static quest databases or repetitive content.

Natural Language Tutorials

Relationship-Building AI

AI That Takes Action

Give AI direct control over game mechanics. Let NPCs spawn enemies, modify environments, or distribute rewards based on context. Ship dynamic gameplay without hardcoding every possibility.

Dynamic World Generation

Generate quests, lore, and world events on the fly. Your game world can create infinite content based on player actions and preferences - no more static quest databases or repetitive content.

PRICING

Price Scales With Your Success

Start free. Pay only when your game succeeds.

Community

Static knowledge base overlay

Free forever

✓ Works with any published game
✓ Zero integration required
✓ Wiki/docs knowledge base
✓ Source references in answers
✓ 100% local, private, offline

Add Your Game

Companion

Real-time game state awareness

5% × reviews × price

One-time payment

Everything in Community, plus:

✓ Real-time player state awareness
✓ Reads location, stats, quests, inventory
✓ Contextual hints, not generic answers
✓ Personalized onboarding & reminders
✓ Adapts to player behavior

Book a Demo

Engine Integration

Full Tryll Engine power

10% × reviews × price

One-time payment

Everything in Companion, plus:

✓ Custom in-game characters
✓ Voice integration (TTS/STT)
✓ AI that takes action in-game
✓ Dynamic world generation
✓ Relationship-building AI with memory
✓ New mechanics: mimics, commentary

Book a Demo

Price Calculator

Tryll scales with your success

Game Price

Custom price: $

Steam Reviews: 10,000

0 10k 20k 100k

Select Tier

One-time price

$5,000

10,000 reviews × $10 × 5%

Estimated game revenue

$5,000,000

Based on 1 review ≈ 50 sales

FAQ

Frequently Asked Questions

Everything you need to know about Tryll Engine

To maintain 90 FPS while running local AI inference, we recommend a GPU like the RTX 4070 Ti Super or equivalent. Tryll automatically selects the most efficient model and quantization for the user's system to ensure smooth gameplay without compromising AI responsiveness.

Our benchmarks show that ~40% of PC gamers can already run Tryll's default models (7B–13B) at acceptable speeds. As open-source models get smaller and faster every month, and gamers continue upgrading hardware, this percentage grows - unlocking scalable deployment without cloud reliance.

Local AI advantages:

Zero cost per player - no API or infrastructure bills
Offline capable - your game keeps working without an internet connection
Private by design - no personal data leaves the device

Cloud-based AI adds legal and UX risks:

Immersion-breaking outages during play sessions
Legal compliance challenges (GDPR, AI Act, export laws)

Tryll's local-first approach is not just cheaper - it's safer, faster, and scalable to millions of players.

Tryll includes a local-first LLM Manager that handles:

Detecting available GPU and RAM
Selecting the optimal model size and quantization (4B-14B)
Auto-updating models as better open-source options become available
Dynamically adjusting context usage by offloading high-level reasoning to the Director agent and storing memory in a local, searchable context database
Falling back to traditional dialogue if the system cannot run models

All of this happens invisibly at runtime - no setup required from players or developers.

Tryll Engine is designed with stability as a top priority. The AI runs in a separate process with robust error handling and automatic recovery mechanisms. If the AI system encounters any issues, it gracefully falls back to default behaviors without affecting core gameplay. Our extensive testing shows no impact on game stability when properly integrated.

Tryll is optimized to have minimal impact on game performance. On recommended hardware (RTX 4070 or equivalent), you can expect a 5-10% performance impact while maintaining smooth gameplay. The AI automatically adjusts its resource usage based on your system's capabilities, ensuring your gaming experience remains fluid and responsive.

You can still run Tryll with a smaller model (like 3B parameters), but the AI quality won't be as good. We find models of 7B+ parameters to be reliable and fun to play with, while 3B models provide a noticeably less engaging experience. We recommend RTX 4070 or equivalent for the best results.

Tryll runs entirely on your device. We don't collect any personal data and have no access to your chat logs. Everything stays local.

Still have questions?

Power Your Game
with LLMs

Start Simple. Go Deeper When Ready.

Tell Us Your Game

We Do the Heavy Lifting

Players Stay Immersed

We Handle the Hard Parts

Zero Infrastructure

Prompts Done For You

Start Without Code

See Tryll Engine in Action

Natural Language Tutorials

Relationship-Building AI

AI That Takes Action

Dynamic World Generation

Natural Language Tutorials

Relationship-Building AI

AI That Takes Action

Dynamic World Generation

Price Scales With Your Success

Community

Companion

Engine Integration

Price Calculator

Frequently Asked Questions

What kind of hardware is needed to run both the game and AI at 90 FPS?

How many gamers can actually run these models?

Why is local AI better than using a cloud API?

How does Tryll manage AI models and inference on the player's machine?

How does using Tryll Engine affect game stability?

How does the use of Tryll Engine affect performance of the game?

What if my hardware isn't powerful enough?

How does Tryll handle my personal data and conversations?

Book a Demo

Power Your Game with LLMs

Start Simple. Go Deeper When Ready.

Tell Us Your Game

We Do the Heavy Lifting

Players Stay Immersed

We Handle the Hard Parts

Zero Infrastructure

Prompts Done For You

Start Without Code

See Tryll Engine in Action

Natural Language Tutorials

Relationship-Building AI

AI That Takes Action

Dynamic World Generation

Natural Language Tutorials

Relationship-Building AI

AI That Takes Action

Dynamic World Generation

Price Scales With Your Success

Community

Companion

Engine Integration

Price Calculator

Frequently Asked Questions

What kind of hardware is needed to run both the game and AI at 90 FPS?

How many gamers can actually run these models?

Why is local AI better than using a cloud API?

How does Tryll manage AI models and inference on the player's machine?

How does using Tryll Engine affect game stability?

How does the use of Tryll Engine affect performance of the game?

What if my hardware isn't powerful enough?

How does Tryll handle my personal data and conversations?

Book a Demo

Power Your Game
with LLMs