Introducing Chat with Your Data, AI Memory & Semantic Search

# Introducing Chat with Your Data, AI Memory & Semantic Search

Today we are shipping a set of capabilities that fundamentally change what you can do inside Vincony: the ability to chat with your own documents and connected workspaces, persistent AI memory that carries context across every conversation, and semantic search that lets you find the right AI model — or a specific image in your asset library — by describing what you need in plain English.

This is not a minor update. Together, these features turn Vincony from a unified AI chat platform into a genuine knowledge workspace. Here is what launched, why it matters, and how to get started.

Chat with Your Data

AI chat is most useful when the model has access to the right information. Until now, every conversation on Vincony started from scratch — you brought the context, the model brought the intelligence. That changes today.

With Chat with your data, you can:

• Paste a document directly into any chat — meeting notes, a research paper, a product brief — and interrogate it instantly.
• Crawl a URL and have Vincony index the page so the model can answer questions grounded in that content.
• Index a public GitHub repository and ask questions about the codebase — architecture, function behaviour, dependency choices — without copying and pasting files.
• Connect Notion, Google Docs, Google Drive, or Slack and have your workspace knowledge available in every chat, automatically kept in sync.

When you ask a question, the model draws on your indexed content and includes source citations in its answer — so you always know where the information came from. Indexing and retrieval run on Vincony's own infrastructure, so none of this consumes your chat credits.

Plan requirements: Pasting documents, crawling URLs, and indexing GitHub repos are available on the Starter plan ($16.99/mo) and above. Connecting Notion, Google Docs, Google Drive, and Slack requires the Pro plan ($24.99/mo) and above. All indexed data is private to your account or workspace — it is never used to train models or shared with other users.

AI Memory

One of the most consistent frustrations with AI assistants is having to re-establish context at the start of every conversation. You explain your tech stack, your writing style preferences, your project goals — and then you close the tab and do it all again tomorrow.

AI Memory fixes this. As you chat, Vincony quietly builds a memory of the things that matter: preferences you have expressed, facts about your projects, constraints you have mentioned. In future conversations, the model draws on this automatically — it remembers that you prefer concise answers, that your project uses a particular framework, that you already explored a certain approach last week.

Alongside memory, you can search your entire conversation history at /os/conversation-search. Type a phrase, a concept, or a question — even if you cannot recall the exact words you used — and Vincony surfaces the threads where that topic came up. Deep links take you straight to the relevant exchange.

Plan requirement: AI Memory and conversation search are available on the Pro plan ($24.99/mo) and above. Your conversation history is stored privately and is never visible to other users.

Semantic Search for AI Models

Vincony gives you access to 800+ AI models across 80+ providers. That breadth is powerful, but it raises an obvious question: how do you find the right model for an unfamiliar task?

The natural-language model finder lets you describe what you need — "a model that is good at translating legal documents into plain English" or "a fast, affordable model for classifying customer support tickets" — and Vincony returns ranked matches with an explanation of why each model fits. Every result page also shows Similar models, so you can compare alternatives side by side without building your own shortlist from scratch.

Semantic model search and the Similar models feature are free on every plan, including the free tier.

Visual Image Search

If you use Vincony to generate or store images, the new visual image search in the Asset Library lets you find them by describing what they show — not by filename or tag. Build an index of your saved images and search for "a minimalist product shot on white background" or "a chart showing upward growth" and the results reflect the actual content of each image.

Saved-prompt search works the same way: describe the kind of prompt you remember saving and Vincony retrieves it, even if you cannot recall the exact phrasing.

Plan requirement: Visual image search and saved-prompt search are available on the Pro plan ($24.99/mo) and above.

No Extra Credits. Your Data Stays Private.

A note on billing: indexing your documents, crawling URLs, syncing connectors, and running semantic retrieval all happen on Vincony's own infrastructure and do not consume your chat credits. You pay for AI model calls, not for the knowledge layer that makes those calls more useful.

All data you bring into Vincony — documents, conversations, connected workspace content — is stored privately under your account. It is not shared across accounts, not visible to other users, and not used for model training.

How to Get Started

• Try semantic model search right now — it is free. Open the model library and type a description of what you need.
• Add your first document — on Starter or above, open /os/knowledge-sources and paste in a document or crawl a URL.
• Connect your workspace — on Pro, link Notion, Google Docs, Drive, or Slack from the same page.
• Enable AI Memory — on Pro, toggle memory on in your account settings and it will begin building from your next conversation.
• Explore visual image search — on Pro, open the Asset Library and build your image index.

Not on the right plan yet? Compare features and pricing at /pricing — upgrading takes under a minute and the new capabilities are available immediately.

---

These features mark a significant step in what Vincony is: not just a place to access AI models, but a workspace that gets smarter with every document you share and every conversation you have. We are building a lot more on top of this foundation — stay tuned.