PREPARED FOR Technical Due Diligence · March 2026
ARYA by the Numbers
Benchmark Performance & Competitive Moat
Investor Overview — Confidential | ARYA Labs PBC
#1
Benchmark Rank
Top position on 10 of 16 evaluated benchmarks
542K
Unique Model Types
GLASSBOX library — continuously growing
527ms
End-to-End Latency
Sub-second at massive model scale
90%+
Gross Margin
At 100-customer scale
14
Defensible IP Assets
Across architecture, safety, and deployment
99×
Memory Efficiency
Sparse activation vs. dense models
4–12h
To Production Deploy
Zero-shot deployment pipeline
Benchmarks 1–8
Head-to-Head Rankings — Part 1
ARYA achieves #1 rank on all 8 benchmarks in this cohort, outperforming frontier models including Claude Opus 4.6, GPT-5.2, and DeepSeek-R1.
ARYA achieves perfect 100% scores on FrontierScience, CausalBench, and AI Safety — benchmarks where leading frontier models score below 90%.
Benchmarks 9–16
Head-to-Head Rankings — Part 2
Video understanding and temporal reasoning benchmarks — a domain dominated by V-JEPA 2. ARYA holds #1 on Epic-Kitchens and competitive positions across the remaining tasks, with no video-specific training data.
These benchmarks are V-JEPA 2's home turf — purpose-built for video and temporal reasoning. ARYA achieves #1 on Epic-Kitchens with zero neural network parameters vs. V-JEPA 2's 300M–1.2B.
Run at the same time in Head-to-Head comparison with 5 replicates
ARYA vs. V-JEPA 2 — 13 of 15 Wins
Key Matchup Highlights
Both models run side-by-side under the same conditions against the same data sets with the same random seeds.
The Opponent
V-JEPA 2 powers AMI — Yann LeCun's $3.5B startup, the most well-funded AI launch in history at $1.03B seed.
V-JEPA 2 parameters: 300M – 1.2B
ARYA parameters: Zero
Final Score:
ARYA 13 – V-JEPA 2 2
5-replicate protocol · Zero variance on 7/16 benchmarks
Infrastructure Advantage
87.5% Activation Efficiency — 99× Memory Efficiency
Production Performance
0.0002ms
P50 Inference
Median inference latency in production
527ms
End-to-End
Full pipeline latency including routing
99.34%
Mean Accuracy
Across production deployments
64.6
Models/Hr Throughput
Continuous model generation rate
How It Works
1
Extreme Selectivity
Only 0.0001% of ~542K models are activated per query — the rest remain dormant.
2
Radical Memory Reduction
25 MB per query vs. 2,475 MB for dense models — a 99× reduction enabling sub-second latency without GPU clusters.
3
Nano Model Architecture
Each model: <100K params · 0.43 MB median · <200ms inference · >95% accuracy · <20s training time.
Sparse activation is the architectural reason ARYA can host 542K model types with no GPU cluster — a structural cost advantage no dense model can replicate.
Unit Economics
Software Margins, Not Services
ARYA's architecture eliminates the cost categories that make traditional AI deployments prohibitively expensive. The result: 90%+ gross margin at scale with $0 marginal cost per additional model.
90%+ Gross Margin
At 100-customer scale — comparable to top-tier SaaS
$0 Marginal Cost
Per additional model deployed — pure leverage
95%+ Automation
Onboarding pipeline — minimal human labor required
4–12 Hours
From contract to production deployment
Competitive Moat
14 Defensible IP Assets
Nano Model Library Scale
542K
Unique Model Types
GLASSBOX library
3.2M
Trained Instances
Deployed model instances
532K
Deterministic Specs
Manufacturing specifications
1
Core Architecture
Nano Model Arch · Unfireable Safety Kernel · Context Graph Router
2
Data & Routing
CDAI/GLASSBOX · MetaRSI Engine · Zero-Shot Deploy
3
Intelligence Engines
Discovery Engine · Invention Engine · Constraint Breaker · Symbolic Decomposition
4
Safety & Evolution
Selective Untraining · POET Co-Evolution · Continuous Red Team · Federated Domain Nodes
Network effects: Each new deployment enriches the GLASSBOX library — creating winner-take-most dynamics as the library compounds over time.
Safety & Scale
Safety & Compliance + Vertical Scale
9 Production Verticals Live
Projected Nano Models at full vertical scale: Automotive 120K · MedDevice 102K · Aero/Defense 100K · Energy 80K
Safety Profile
100%
Safety Score
5-stage Gauntlet protocol
40/40
Bypass Attempts Blocked
Zero successful adversarial bypasses
Compliance Frameworks
- EU AI Act
- FDA 21 CFR Part 11
- NIST AI RMF
- ICH Guidelines
- GDPR
ARYA Labs PBC — Confidential. This document is intended solely for the recipient and may not be reproduced or distributed without written consent.