AI Ecosystem Statistics 2026
As of September 2026, the AI ecosystem has 4,285,300 indexed assets across 2,336,700 models, 223,500 agents, 53,600 tools, 62,100 MCP servers, 712,500 datasets, 821,800 spaces, and 62,200 containers. The average trust score is 45/100. Data from Nerq, based on 13 cross-referenced sources updated daily.
4,285,300
Total Assets
2,336,700
Models
62,100
MCP Servers
45
Avg Trust Score
Asset Breakdown
| HuggingFace Models | 2,336,700 |
| HuggingFace Spaces | 821,800 |
| HuggingFace Datasets | 712,500 |
| AI Agents | 223,500 |
| AI Tools | 53,600 |
| Docker Containers | 62,200 |
| MCP Servers | 62,100 |
Trust Grade Distribution
| Grade | Count |
|---|---|
| D | 3,963,200 |
| E | 336,000 |
| C | 14,000 |
| C- | 2,700 |
| B | 1,200 |
| F | 1,200 |
| B- | 1,000 |
Source: Nerq analysis of 4,285,300 AI assets from GitHub, HuggingFace, npm, PyPI, Docker Hub, and 8 other sources. Updated September 2026. API access