AI game platform due diligence: what investors actually look at in 2026

Venture capital in AI game platforms reached $180M in 2026, but not all platforms deserve investor attention. This guide reveals what serious investors examine during due diligence—and why many AI game companies fail to pass scrutiny.

The quality-velocity trap

A recent analysis by Dre Dyson, a VC specializing in game startups, identified a critical pattern: platforms optimizing for raw shipping velocity over technical quality consistently trade at 40-60% lower revenue multiples.

"Steam's 32% supply growth with only 19% quality growth is the public market equivalent of a down round. The teams that survive are the ones that treat AI like nuclear power—incredible energy density, but you need containment buildings." — Dre Dyson, VC

Orbit Arcade AI game generation pipeline

What this means for AI game platforms

| Metric | Healthy Platform | Red Flag | | --- | --- | --- | | Generation success rate | 85%+ | Below 70% | | First-attempt playability | 80%+ | Below 60% | | User retention (week 4) | 40%+ | Below 20% | | Creator revenue share | Transparent | Undisclosed | | AI cost per player | Known and optimized | Unknown |

The four due diligence questions

Investors who understand AI game platforms ask these questions first:

1. "What is your quality-to-velocity ratio?"

The Quality-to-velocity Ratio (QVR) measures whether speed creates value or destroys it:

QVR = (Mean_Time_To_Resolution * Deploy_Frequency) / 
      (Defect_Rate * Churn_Rate_Attributed_To_Quality)

Target for Series A: QVR > 2.5
Red flag: QVR < 1.0 with high AI code generation %

What investors see: Platforms with high generation volume but low retention have collapsing QVR. The marginal games flood the market as "AI slop"—technically functional but commercially hollow.

2. "Walk me through your AI cost at scale"

Orbit Arcade AI infrastructure

| Scale | Cost per Player | Total AI Cost | | --- | --- | --- | | 1,000 MAU | $0.12 | $120 | | 10,000 MAU | $0.08 | $800 | | 100,000 MAU | $0.05 | $5,000 | | 1,000,000 MAU | $0.03 | $30,000 |

Red flag: "We'll figure that out later." Teams that cannot articulate AI costs at scale have not thought through the business model.

3. "What is your fallback when the model fails?"

Platforms dependent on a single AI provider face significant risk. API prices change, models get deprecated, rate limits shift.

What investors look for:

Multi-model routing (not single-provider dependency)
Local fallback strategies
Cost optimization through caching and tiering

4. "Show me your creator flywheel"

Orbit Arcade creator ecosystem

The chicken-and-egg problem: you need players to attract creators, and creators to attract players.

| Flywheel Stage | Healthy Signal | Concerning Signal | | --- | --- | --- | | Creator supply | Growing organically | Requires constant payouts | | Player retention | Week-4 > 30% | Week-4 < 15% | | Content quality | Improving over time | Declining with volume | | Revenue per creator | Growing | Flat or declining |

Astrocade: a case study in due diligence concerns

Naavik's analysis of Astrocade's $56M raise identified several concerns investors should note:

Orbit Arcade portfolio showcase

The friction points

Publishing latency: Games take up to a week to surface in feeds—contradicting "instant creation" claims
Quality concerns: "The platform feels like an aggregation of game demos, often with clunky controls and limited progression depth"
Retention questions: "Can prompt-based generation create experiences able to meaningfully attract and retain players?"

The engagement questions

Chris Heatherly's LinkedIn analysis raised pointed questions about Astrocade's metrics:

"Big numbers are great, but real momentum is retention + active creators + organic growth. That's the part I'd watch closely."

Specific concerns raised:

Social media followers don't match claimed user numbers
Google Trends shows search traffic flat since funding announcement
Creator payouts may incentivize "farming" rather than genuine creation
Discord engagement appears lower than claimed metrics suggest

The competitive moat question

"Unless Astrocade is truly able to unlock a legion of new game makers and have its AI system get better at creating novel, fun, and deeper experiences, it will likely face strong competition." — Naavik

Aippy: the credit system problem

App Store reviews reveal consistent issues:

"I was having loads of fun creating my own game but then out of nowhere it just says you're out of credits. I was devastated." — Aippy user

"The AI isn't very good. I put a huge prompt into it and explained everything in perfect detail but the game was nothing like how I described it." — Aippy user

Due diligence signals:

Credit system creates friction for creators
AI accuracy complaints suggest generation quality issues
Bug reports indicate stability concerns

Loopit: the depth limitation

AppCandyBox's review identified structural limitations:

"The same simplicity that makes Loopit approachable can make it feel shallow once the novelty wears off... Loopit sells spontaneity better than depth."

Key concerns:

Daily generation limits frustrate serious creators
AI consistency issues ("breaking spriting systems")
Accusations of inflated engagement metrics
Content moderation challenges with "inappropriate" content

Rezona: the consistency problem

Google Play reviews highlight generation reliability issues:

"It can build a game, but not anything serious. It is really slow and inconsistent, you have to go through multiple iterations just to switch an image." — Rezona user

"Adding limits was a huge mistake... having unlimited usage was already good enough so adding limits and tokens were completely unnecessary." — Rezona user

Due diligence signals:

Slow and inconsistent generation under load
AI overwriting creator work without permission
Monetization changes frustrating existing users
Limited ability to update published games

What Orbit Arcade does differently

The platforms that pass investor scrutiny share common traits:

| Factor | Orbit Arcade | Competitor Pattern | | --- | --- | --- | | Generation consistency | 94% first-attempt success | 58-76% typical | | Cost transparency | Published economics | Undisclosed | | Creator flywheel | Feed-native discovery | Gallery-dependent | | Iteration support | Comment-to-fix workflow | One-shot generation | | AI architecture | Multi-model routing | Single-provider |

"The future isn't just finding your next game. It's experiencing it instantly." — Christian Frithiof, IGDB founder

Orbit Arcade aligns with this vision: creation, discovery, and iteration in one browser-first loop.

Investor checklist

Before investing in any AI game platform, verify:

| Check | How to Verify | Red Flag | | --- | --- | --- | | Generation quality | Test with 50 standardized prompts | Success rate < 70% | | Retention metrics | Request week-1 and week-4 data | Week-4 < 20% | | AI cost structure | Ask for cost-per-player at scale | Cannot articulate | | Creator economics | Verify revenue share model | Undisclosed or unfavorable | | Competitive moat | Assess distribution advantage | Gallery-only | | Team technical depth | Interview CTO on architecture | Cannot explain AI fallbacks |

The quality premium

Platforms that invest in generation quality command premium valuations:

| Quality Level | Valuation Multiple | Investor Interest | | --- | --- | --- | | High (90%+ success) | 15-25x revenue | Strong | | Medium (75-89%) | 8-15x revenue | Moderate | | Low (< 75%) | 3-8x revenue | Limited |

The market has learned to discount "feature factory" output. Quality compounds; volume without quality destroys value.

See Orbit Arcade in action

Experience the difference quality generation makes:

Browse free online games — Play creator-made HTML5 games instantly
Use the AI Game Maker — Create your own playable in seconds
View the product tour — See the full platform workflow