Technology & Utility
20%Protocol architecture, innovation, scalability, security, decentralisation, developer activity and real-world usefulness.
The Crypto AI Gem Score™ methodology
Every asset we track is evaluated by an AI research model against six criteria. The model never decides the final number: it scores each criterion from 0–100, and our backend applies fixed weights to compute the overall score. That keeps results consistent, auditable and comparable across thousands of assets. Current scoring version: v1.0.
The six criteria
Team & Execution
15%Founder and core developer track record, transparency, delivery against roadmap, governance and institutional credibility.
Tokenomics
20%Circulating supply, dilution, unlock schedules, insider and VC allocation, emissions, vesting and token utility.
Adoption & Ecosystem
15%Active users, transactions, TVL, applications, integrations, exchange availability, community growth and network revenue.
Funding & Partnerships
10%Investor quality, capital raised, institutional backing and confirmed partnerships — announcements alone do not count.
Growth Potential & Risk
20%Upside versus downside: valuation, sector growth, narrative, liquidity, regulatory, unlock, competition and execution risk.
How the score is calculated
Overall score = the sum of each criterion score multiplied by its weight, rounded to the nearest whole number. Weights always add up to 100%, so a coin that scores 80 on every criterion receives exactly 80 overall.
Technology & Utility × 20% Team & Execution × 15% Tokenomics × 20% Adoption & Ecosystem × 15% Funding & Partnerships × 10% Growth Potential & Risk × 20%
Rating bands
- ELITE GEM95–100
- STRONG GEM90–94
- HIGH POTENTIAL85–89
- PROMISING80–84
- ON WATCH70–79
- SPECULATIVE60–69
- HIGH RISK0–59
Confidence and data gaps
Each analysis carries a confidence level from 0–100 reflecting how much verifiable evidence the model found. When information is missing — an anonymous team, no audited contracts, undisclosed funding — the model records the gap explicitly and lowers both the relevant criterion score and confidence rather than guessing. Every coin page lists the data gaps behind its score.
Data sources and refresh cadence
Market data (price, market cap, volume, supply, all-time highs) comes from public market data providers and refreshes continuously. Qualitative research draws on project documentation, developer activity, community metrics and exchange listings available at analysis time. Scores are re-generated on a rolling schedule and stored append-only, so the full score history for every asset is preserved.
Limitations
AI research can be incomplete or wrong. Scores are a starting point for your own research, not a recommendation to buy or sell anything. Nothing on CryptoAIGem is financial advice.