Skip to main content
Many teams adopt Omnifact after using consumer tools like ChatGPT or Microsoft Copilot. While the core chat experience is immediately familiar, Omnifact is built around enterprise data privacy, multi-model choice, and German/EU works council compliance. This guide answers the most common “Does it work like ChatGPT?” questions and clearly lists what Omnifact does (and does not) do today.

​
ChatGPT vs. Omnifact at a Glance

​
What Omnifact Doesn’t Do (and What to Use Instead)

1. Automatic memory across conversations

Does Omnifact automatically remember details from my previous chats?No. Each conversation in Omnifact is isolated so confidential details from one client, HR case, or project never leak into unrelated chats.What to use instead:
  • For your role and writing style: Go to Account Settings > Personalization and fill in About (your role and department) and Response Preference Description (how you want the agent to write). The agent applies these to all your chats.
  • For recurring project or team context: Create a Space and add the background context to Chat Instructions or upload reference documents to Uploaded Files.

2. Sharing individual chat threads

Can I send a colleague a link to my chat conversation?No. Every chat conversation (including conversations you have inside a shared Space) is private to your user account. Neither colleagues nor organization administrators can browse your chat history.What to use instead:
  • To work with the same AI setup and documents: Share a Space with your colleagues or user group. Everyone in the Space talks to the same configured agent and searches the same Knowledge Base in their own private chat threads.
  • To share a result: Click the Copy icon under any AI response, or ask the agent to generate a formatted Word, Excel, or PowerPoint file using Document Creation.

3. Importing ChatGPT history or Custom GPTs

Can I import my ChatGPT export file or migrate Custom GPTs automatically?No. Omnifact does not support importing chat archives or Custom GPT packages from ChatGPT.How to migrate a Custom GPT in 2 minutes:
  1. Click Spaces in the left navigation bar and click New Space.
  2. Copy the system instructions from your Custom GPT into the Space’s Chat Instructions field.
  3. Add your conversation starters under Example Prompts.
  4. Upload any knowledge files in Space Settings > Knowledge > Uploaded Files (or link a folder via Connected Sources).

4. Native mobile app (iOS / Android)

Is there a dedicated Omnifact mobile app?No. Omnifact does not currently offer a native iOS or Android app. You can access Omnifact on smartphones and tablets by opening https://chat.omnifact.ai in your mobile web browser.
Can we upload a custom dictionary of terms for the Privacy Filter to ignore or mask?No. There is no organization-wide allowlist or custom entity list. To share specific terms with the model without masking:
  • Click any masked word in your sent message to reveal it for the rest of that chat
  • Wrap a term in double percent signs (%%term%%) when typing to exclude it inline
  • If your administrator has granted your group the Privacy Filter Exemption, toggle on Privacy Filter Bypass under + > Privacy Controls for that chat

​
Most Frequently Asked Questions

  • Covered: Typed chat messages, extracted text from attached documents, retrieved Space knowledge chunks, and text output (stdout, stderr, return values, and user attachment filenames) from sandboxed code execution.
  • Not covered: Visual pixels inside attached images, spoken audio in audio files or voice dictation, external MCP server tool calls, and requests to the Public API or AI Gateway.
See the full coverage matrix and limitations in How the Privacy Filter Protects Your Data.
  • The €9/user monthly AI credit is included in the standard €25/user seat price and is pooled into one shared monthly budget for your organization.
  • Unused budget does not roll over to the next month.
  • All model tiers (Base, Standard, and Premium) draw from the budget while credits remain. When the organization budget runs out, Base Tier models stay available as a fallback for the rest of the cycle.
  • Administrators can set a Per-user spending limit so a single heavy user cannot drain the shared pool, and can top up credits immediately via Add More Credits.
See AI Credits and Usage Limits for full details.
  • Region flags: A model with an 🇪🇺 EU flag processes prompts in the European Union; a model with a 🇺🇸 US flag processes prompts in the United States. Administrators can restrict the workspace or specific groups to EU-hosted models only.
  • No training: Provider agreements across all models exclude using your data for model training.
  • Zero Data Retention (ZDR): In place for Mistral AI, Anthropic (EU), and Google, and not yet in place for Azure OpenAI.
See Data Residency, Retention & Compliance for the complete provider table and Trust Center links.
Document Creation and Code Execution must be enabled for your user group by a team administrator. Ask your administrator to go to Team Settings > Groups, open your group (or the Everyone group), and enable Execute Code (incl. Document Creation) under the Permissions tab.
Documents uploaded to a Space’s Uploaded Files are split into searchable text chunks, so the AI only retrieves short excerpts and never sees the whole file or its formatting at once. To have the agent fill out or edit a template while preserving its structure, attach the file directly in the chat input (with Document Creation enabled). See Attachments vs. Uploaded Files.