Nerq found 6,735 MCP servers for llm & model serving, ranked below by GitHub stars and weighted compliance score across 52 global AI jurisdictions. The top-ranked MCP server for llm & model serving is 54yyyu/zotero-mcp with 1,461 stars and a trust score of 82/100. MCP servers for interacting with large language models, model serving, and AI inference APIs.
Last updated: August 01, 2026 | 6,735 servers in this category | Data from Nerq| # | Name | Stars | Trust | Author | License | Description |
|---|---|---|---|---|---|---|
| 1 | 54yyyu/zotero-mcp | 1,461 | A 82 | 54yyyu | MIT | |
| 2 | gatekit | 37 | B 75 | gatekit-ai | Apache-2.0 | |
| 3 | Node.js Debugger | 303 | B+ 75 | https://github.com/workbackai/mcp-nodejs-debugger | N/A | |
| 4 | Toolbox | 30 | B+ 75 | https://github.com/ai-zerolab/mcp-toolbox | N/A | |
| 5 | Supabase | 4 | B+ 75 | https://github.com/alexander-zuev/supabase-mcp-server | N/A | |
| 6 | jalpp/chessagine-mcp | 18 | B 75 | jalpp | MIT | |
| 7 | DareDev256/fcpxml-mcp-server | 10 | B 74 | DareDev256 | MIT | |
| 8 | Deep Code Reasoning | 106 | B 74 | https://github.com/haasonsaas/deep-code-reasoning-mcp | N/A | |
| 9 | triage-ai | 0 | B 74 | wyman101 | MIT | |
| 10 | ableton-mcp | 0 | B 74 | jpoindexter | MIT | |
| 11 | tomaspavlin/rohlik-mcp | 82 | B 73 | tomaspavlin | N/A | |
| 12 | rangta10/kali-mcp-server | 3 | B 73 | rangta10 | NOASSERTION | |
| 13 | cpal | 1 | B 73 | tobert | MIT | |
| 14 | aashari/mcp-server-atlassian-bitbucket | 115 | B 73 | aashari | N/A | |
| 15 | Azure/aks-mcp | 115 | B 73 | Azure | N/A | |
| 16 | takashiishida/arxiv-latex-mcp | 100 | B 73 | takashiishida | N/A | |
| 17 | deliberate-reasoning-engine | 4 | B 73 | haasonsaas | MIT | |
| 18 | k3s-mcp-server | 1 | B 73 | ry-ops | MIT | |
| 19 | BerriAI/litellm | 36,444 | B 73 | BerriAI | NOASSERTION | |
| 20 | ConnorBritain/mssql-mcp-server | 7 | B 73 | ConnorBritain | MIT |
MCP servers for interacting with large language models, model serving, and AI inference APIs. Nerq tracks 6,735 MCP servers in the llm & model serving category, each assessed against 52 global AI regulations including the EU AI Act, US state AI laws (California SB53, Colorado SB205), UK AI Bill, and more. Compliance scores are weighted by jurisdiction penalty severity — jurisdictions with higher fines (like the EU AI Act at up to €35M) have more impact on the score than voluntary frameworks.
All data is updated regularly. Use the Nerq API or MCP Server for programmatic access.