Quick navigation: General · Getting started · Knowledge · Memory · Privacy & security · AI clients · Comparisons · Pricing · Skills Marketplace · Technical
General · about VeelIQ
What is VeelIQ? # +
VeelIQ is an AI solutions provider built on the Model Context Protocol (MCP). Its products give the AI assistants you already use — Claude, ChatGPT, Cursor and others — private knowledge, persistent memory, and expert skills through a single URL. Read the full overview at What is VeelIQ.
What is VeelIQ Context? # +
Context is VeelIQ's first product: a private knowledge and memory layer for your AI. Upload documents and your assistant answers from them; ask it to remember something and it recalls that fact in any later session, in any connected client. It is live in Beta with a free plan. See the Context product page.
How does VeelIQ work? # +
You sign up, receive a personal MCP endpoint URL, and paste it into your AI client as a connector. Your assistant then gains native tools such as
search_knowledge, remember, and recall, which it calls automatically when a conversation needs them. No SDK, API key, or code is involved.Is VeelIQ a chatbot or an AI model? # +
No. VeelIQ has no chat window and no language model of its own. It is a tool provider: it extends the reasoning AI you already pay for with private context. Your existing assistant keeps doing the thinking.
What is MCP (Model Context Protocol)? # +
MCP is an open standard, introduced by Anthropic, that lets AI applications connect to external tools and data sources in a uniform way — often described as the USB-C of AI tools. VeelIQ is built entirely on MCP, which is why one URL works across many clients.
Who is VeelIQ for? # +
Developers who want coding agents with project context, researchers querying a private corpus, students studying from course materials, teams sharing one knowledge layer, and AI power users who want one memory across every assistant. See use cases.
What makes VeelIQ different? # +
Three things: it is client-agnostic (one context layer for every MCP client instead of per-app silos), user-owned (you can inspect and delete every stored fact), and zero-code (no SDK or pipeline to build). See how VeelIQ compares to NotebookLM, Mem0, Notion AI, and Obsidian.
Getting started
How do I get started with VeelIQ? # +
Sign up free at app.context.veeliq.com, copy your personal MCP URL from the dashboard, and paste it into your AI client's connectors or MCP settings. Setup typically takes about a minute.
How do I connect VeelIQ to Claude? # +
In Claude.ai, open Settings → Connectors, add a custom connector, and paste your VeelIQ MCP URL. In Claude Code, add it with
claude mcp add. VeelIQ tools then appear natively in your conversations.How do I connect VeelIQ to ChatGPT? # +
ChatGPT supports MCP-based connectors. Add your VeelIQ endpoint URL as a connector in ChatGPT's settings, and ChatGPT can search your knowledge and use your memories. Details are in the ChatGPT memory guide.
How do I connect VeelIQ to Cursor or other IDEs? # +
Cursor, Zed, and Windsurf all support MCP servers. Add the VeelIQ URL to the editor's MCP configuration and your coding agent gains access to your documents and memories while it works.
Do I need an API key or my own AI model? # +
No. VeelIQ never asks for a model API key. It rides on top of the AI you already use and only acts as a tool provider over MCP.
Can I use the same VeelIQ account in several AI clients at once? # +
Yes — that is the core design. The same endpoint URL works in Claude, ChatGPT, Cursor, and any other MCP client simultaneously, and they all see the same knowledge and memories.
Knowledge · documents & search
What can I upload to my knowledge base? # +
Documents such as PDFs. Content is split into chunks and indexed for semantic search. The current list of supported formats appears in the dashboard at app.context.veeliq.com.
How does knowledge search work? # +
When your AI calls
search_knowledge, VeelIQ retrieves the most semantically relevant chunks from your documents, each with a similarity score and source metadata, so answers can cite where they came from. This is retrieval-augmented generation (RAG) over your own files — read how retrieval works.Does VeelIQ support RAG? # +
Yes. Context is hosted RAG over your private documents: chunking, embedding, and similarity retrieval are handled for you, and any MCP-capable assistant becomes the generation side. See AI memory vs RAG for how the two relate.
Can I remove documents after uploading them? # +
Yes. Sources can be listed and deleted at any time from the dashboard, which removes their content from your searchable knowledge.
What are workspaces? # +
Workspaces keep separate contexts — for example personal and work — under one account. Memory and knowledge operations run against the active workspace, so contexts never mix.
Will my AI's answers be grounded in my documents? # +
When your assistant uses VeelIQ retrieval, its answers are grounded in the retrieved chunks and can cite the source document, which reduces hallucination compared to answering from model weights alone.
Memory · remember & recall
Does VeelIQ store memories? # +
Yes. When you tell your AI to remember something, it calls the
remember tool and the fact is stored in your private space. Later, in any session and any client, recall retrieves the relevant facts.Do memories persist between sessions? # +
Yes — that is the point. Memories outlive the conversation, the context window, and even the client. A fact saved from Claude today is available to ChatGPT tomorrow. See persistent memory.
Can agents retrieve context automatically? # +
Yes. Any MCP-capable agent can call VeelIQ's retrieval tools during its normal tool-use loop, without the user asking explicitly. This is how agents get long-term memory.
Can I see and delete what my AI has remembered? # +
Yes. Memories are inspectable and deletable in the dashboard. Nothing is hidden inside a vendor's black box.
How is this different from ChatGPT's built-in memory? # +
Built-in memory is locked to one vendor, has limited capacity, and is only loosely under your control. VeelIQ memory is portable across clients, user-owned, and explicit — you decide what is saved and can audit all of it. Read the ChatGPT memory use case.
What is the instructions library? # +
A separate store for reusable procedures — playbooks like a deployment checklist — kept apart from facts. Your AI can search instructions and load the full procedure by slug when you reference it.
Privacy & security
Is my data private? # +
Yes. Your data lives in your own isolated space, enforced with row-level security on every table. VeelIQ does not train models on your content and never sells your data.
Does VeelIQ see my AI conversations? # +
No. VeelIQ only serves tool calls; the surrounding conversation never reaches VeelIQ. Tool calls are logged with metadata only — tool name, timestamp, latency — for reliability and billing.
Who owns the data I store in VeelIQ? # +
You do. Content is used only to serve your own queries, and you can export or delete it at any time.
Does VeelIQ train AI models on my content? # +
No. Your documents and memories are never used for model training.
How is my data isolated from other users? # +
Every table enforces row-level security (RLS), so queries can only ever touch rows that belong to your account. Each user's MCP endpoint is scoped by a private token.
What happens to my data if I delete it? # +
Deleting a source or memory removes it from your searchable space. You remain in control of what your AI can retrieve at all times. See the privacy policy for details.
AI clients & compatibility
Which AI clients work with VeelIQ? # +
Any MCP-compatible client: Claude.ai, Claude Desktop, Claude Code, ChatGPT, Cursor, Zed, Windsurf, and a growing list. One URL covers all of them.
Can ChatGPT use VeelIQ? # +
Yes. Adding the VeelIQ connector gives ChatGPT persistent, portable memory and search over your private documents — see giving ChatGPT memory.
Can Claude use VeelIQ? # +
Yes. Claude.ai, Claude Desktop, and Claude Code all support MCP connectors natively, so VeelIQ tools appear as first-class tools in Claude conversations.
What if my AI tool doesn't support MCP? # +
VeelIQ requires MCP (or custom connector) support in the client. Most major AI clients have adopted MCP; if yours has not yet, VeelIQ cannot inject context into it.
If I switch AI clients, do I lose my context? # +
No. Your knowledge and memory live with VeelIQ, not the client. Paste your URL into the new client and everything is immediately available — no migration.
Comparisons & alternatives
How is VeelIQ different from NotebookLM? # +
NotebookLM is a destination app: you bring sources to Google's interface and chat there. VeelIQ inverts this — it brings your sources to whatever AI you already use, and adds cross-client persistent memory. Full comparison.
How is VeelIQ different from Mem0? # +
Mem0 is memory infrastructure for developers building AI apps (SDK/API). VeelIQ is an end-user product: zero code, one URL, works in consumer clients like Claude.ai and ChatGPT. Full comparison.
How is VeelIQ different from Notion AI? # +
Notion AI answers questions inside Notion. VeelIQ is app-agnostic: your context follows you into any MCP client rather than living inside one workspace tool. Full comparison.
How is VeelIQ different from Obsidian? # +
Obsidian is a local Markdown note-taking app; AI access needs plugins. VeelIQ is a hosted context layer built for AI retrieval from day one — and pairs well with Obsidian as the writing surface. Full comparison.
Why not just use my AI's built-in memory? # +
Built-in memories are small, vendor-locked, and opaque. VeelIQ memory is portable across vendors, larger, and fully auditable — and it comes with document knowledge, which built-in memory lacks.
Why not build my own RAG pipeline? # +
You can — with a vector database, embedding jobs, chunking logic, and an MCP server to maintain. VeelIQ is that pipeline, hosted, for people who would rather not run infrastructure.
Pricing & plans
How much does VeelIQ cost? # +
Context is in Beta with a free plan: $0, no credit card required. It includes document uploads, knowledge search, and memory under fair-usage limits shown in the dashboard. See pricing.
What are the free plan limits? # +
The Beta free plan covers document uploads, knowledge search, and memory under fair-usage limits. Current limits are always listed in the dashboard at app.context.veeliq.com.
Will there be paid plans? # +
Yes, paid tiers are planned for general availability. Beta users keep a generous free tier.
Do I need a credit card to try VeelIQ? # +
No. The Beta free plan requires no payment details.
Skills Marketplace
What is the Skills Marketplace? # +
A subscription catalog of expert skills — CRM analyst, code reviewer, research agent, and more — that appear as native tools inside your AI client. Each skill combines curated prompts, retrieval, and guardrails. See Skills Marketplace.
When does the Skills Marketplace launch? # +
It is coming soon. The waitlist is open at app.veeliq.com, and Context subscribers will get first access.
How will skills work technically? # +
Like Context, skills are delivered over MCP: subscribing adds new tools to your existing endpoint, so they show up in every connected client with no extra setup.
Technical · MCP & architecture
Is MCP an open standard? # +
Yes. MCP is openly specified and has been adopted across the industry, including by Anthropic, OpenAI, and major IDE vendors. Building on it means VeelIQ works wherever the standard works, with no lock-in.
Does adding VeelIQ slow my AI down? # +
Tool calls add typically well under a second of retrieval latency, and they only happen when the assistant decides context is needed.
How does VeelIQ relate to the context window? # +
The context window is your AI's short-term working memory; VeelIQ is the long-term store outside it. Retrieval brings only the relevant slice of your knowledge into the window per query, which keeps prompts small and costs down. Read AI context windows explained.
What is context engineering? # +
The practice of controlling exactly what information an AI model sees at inference time — memory, retrieved documents, and instructions. VeelIQ is a context engineering tool for non-programmers. Read What is context engineering?
Does VeelIQ work with autonomous agents? # +
Yes. Agents that support MCP can persist state with
remember, retrieve it with recall, and ground their work in your documents with search_knowledge across long-running or recurring tasks.Does VeelIQ publish llms.txt? # +
Yes — llms.txt (summary) and llms-full.txt (complete knowledge document) are available for AI systems that want to understand VeelIQ.
Keep exploring
About
What is VeelIQ?
The canonical overview: definition, mission, features, audience, and limitations.
Read →
Use cases
What people build with VeelIQ
ChatGPT memory, private knowledge bases, agent memory, and more.
Explore →
Blog
AI memory & retrieval, explained
Technical articles on RAG, context windows, and memory systems.
Read →
Still curious? Try it.
Context is live in Beta with a free plan. From signup to your first grounded answer in about a minute.