AnythingLLM - Hardened Self-Hosted Private ChatGPT for Documents (RAG)
This is a repackaged open source software product wherein additional charges apply for hardening, security configuration, and setup support. AnythingLLM is a self-hosted ChatGPT-style workspace for chatting with your documents using any LLM provider, with built-in RAG, AI agents and vector search in a single hardened c...
- Category
- ML Solutions
- Registry category
- Cybersecurity & ITIT Operations & Infrastructure
- Vendor
- Lynxroute
- Location
- Not disclosed
- Reviews
- Not rated (0 reviews on the listing)
- Listed on
- AWS Marketplace
- Pricing
- 5 plans listed
Each fact carries one Provenance score: the average of its Disclosure and Evidence scores, the same rule the page uses for the overall score. Click any score to see the math behind it.
Vendor Provenance
Who actually operates this agent.
Build Provenance
What the agent is built on.
Reach Provenance
How the agent is distributed and adopted.
Data Provenance
What the vendor discloses about data handling.
Compliance Provenance
What standards the vendor meets, and who verified them.
What the publisher says
This is a repackaged open source software product wherein additional charges apply for hardening, security configuration, and setup support.
WHAT IS ANYTHINGLLM
AnythingLLM is an open-source, self-hosted ChatGPT-style application that turns any document collection into a private, queryable workspace. The server is a Node.js application that ingests PDFs, DOCX, HTML, plain text, audio transcripts, and code, chunks and embeds the content into a vector database, and serves a multi-user web UI for chat and AI agents. It supports any LLM provider through a single switch in the admin UI - OpenAI, Anthropic Claude, AWS Bedrock, Azure OpenAI, Google Gemini, Cohere, Mistral, Groq, OpenRouter, Ollama, LM Studio, vLLM, LiteLLM, and many more. AI agents can browse the web, run SQL, call tools, and execute multi-step workflows. Operators get team workspaces with per-workspace document scope, role-based access, public chat embeds, REST API for programmatic access, an OpenAI-compatible developer API, and a built-in admin UI. Persists users, workspaces, chats, documents, and vectors in embedded SQLite and LanceDB - no external database required. MIT license, no vendor lock-in.
Show the rest of the publisher’s description (23 more lines)
WHAT THIS AMI ADDS
Security hardening:
- Admin user with unique password (EC2 Instance ID) generated at first boot - never baked into the AMI
- JWT signing key, encryption signing key, and salt (>=32 chars each) generated at first boot
- AnythingLLM container bound to 127.0.0.1:3001 only - reachable only through the Nginx reverse proxy with TLS
- Multi-user mode enabled at first boot - the bootstrap auth window closes immediately after admin creation
- No provider API keys baked in - operator configures OpenAI, Anthropic, Bedrock, Ollama, etc. in the admin UI after login
- Anonymous telemetry disabled by default
- Nginx reverse proxy with TLS, HTTP-to-HTTPS redirect, WebSocket support for streaming chat, security headers
- UFW firewall pre-configured - only TCP 22, 80, 443 are exposed
- fail2ban, AppArmor
- CVE scan - every image is scanned for vulnerabilities before release
Out of the box, with no external services:
- Embedded LanceDB vector database - file-based, no separate container or network call
- Native embedder (Xenova/all-MiniLM-L6-v2) running inside the container - no external embedding API key required for RAG
- SQLite storage for users, workspaces, chats - no external database to provision
OS hardening (CIS Level 1):
- CIS Ubuntu 24.04 LTS Level 1 benchmark applied via ansible-lockdown
- auditd, SSH hardening, kernel hardening, IMDSv2 enforced
Compliance artifacts:
- SBOM - CycloneDX 1.6 at /etc/lynxroute/sbom.json
- CIS Conformance Report at /etc/lynxroute/cis-report.html
- CIS Tailored Profile at /usr/share/doc/lynxroute/CIS_TAILORED_PROFILE.md
Plans and pricing
5 plans listed
Linked repositories
- Mintplex-Labs/anything-llm (opens in a new tab)SettleTop scannedScanned Nov 13, 2026 · 27dd2b8
We read every file in this repo and every open source part it pulls in. Here is what turned up.
What we found514things to look at in their own code0 serious, 503 high. Most are lower risk.
258known security bugs in the open source it reuses7 critical, 116 high. Public bugs in the 1,733 outside parts it depends on.
Noneleaked passwords or keys in the codeNothing secret was left in plain sight.
0bits of encryption a future quantum computer could breakOut of 0 encryption uses we spotted. Common today, worth tracking.
CriticalHighMediumLowWhat is inside it1MCP servers0Tools0Skills192Model calls263Data stores64Vector stores383Web services1,733Open source partsRecords we have on fileList of software partsNot collectedList of encryption usedNot collectedList of AI partsAI inventoryWhich bugs actually applyNot collectedMap of what depends on whatDependency graphHow it is deployedOBOMServices and connectionsConnectionsScan attestationNot collectedWhat our scan read in this code. It is not a safety rating.
A linked repository is not a claim that the publisher wrote the code; the registry computes that relationship privately and does not publish it.
Publisher resources
Sources
- Marketplace listingAWS Marketplace (opens in a new tab)
- Legal linkStandardEula (opens in a new tab)
- Company recordOutside company data, matched on lynxroute.com
Reading the score
Disclosure · how many sources corroborate it
Evidence · what kind of evidence backs it
The SettleTop steps above are possible but not automated yet, so today they sit open and a perfect 10 is out of reach. Open information like firmographics and reach has fewer steps to climb, so it sits a little higher for the same visible evidence.
Information shown here reflects the best available information at the time of assessment and may be incomplete, outdated, or subject to change. It should be independently verified before being relied upon for any decision. SettleTop is not liable for any errors, omissions, or decisions made in reliance on this information.
Have information about this agent?
Send it to us for consideration
Spot something missing, wrong, or out of date? Share what you know. Every note reaches our team. It does not change the passport by itself: we review it first.
