Search iThub

Search tools, categories and collections. Type at least two characters.

Explore all tools
↑ ↓ to explore resultsEnter to open · Esc to close

AI Assistants & Chat

Chat interfaces and AI assistants you can run locally or self-host, connected to open or commercial language models.

Choosing guide

41 results in AI Assistants & Chat

Active Pro listings from this category appear first. Your sort applies within Pro and Free results.

Advertisements
Advertisements
Focused collections

Shortlists for AI Assistants & Chat

Explore collections with tools relevant to this category, grouped by deployment, source model or workload.

Self-hosted IT tools →

34 matching tools from this category

Open-source IT tools →

38 matching tools from this category

How to choose ai assistants & chat tools

This category covers chat interfaces and personal assistants that you install on a desktop or host on your own server, then connect to local models or to commercial model APIs. Typical features include conversation history, document upload with retrieval, multi-user access, and connections to external tools through the Model Context Protocol. It differs from agent frameworks because the product is a usable front end out of the box, not a library for building one.

Read the guide

Who uses these tools

  • IT teams offering private chat to employees
  • Developers testing several models side by side
  • Privacy-conscious users running models on laptops

Capabilities to compare

  • Support for local and remote model backends
  • Chat with uploaded files and connected sources
  • User accounts, roles and conversation sharing
  • Tool use through MCP servers and plugins
  • Persistent memory and scheduled assistant routines

Operating considerations

  • Hardware needed if models run on the same machine
  • Where chat logs and uploaded documents are stored
  • Permission-aware retrieval for sensitive company sources
  • Per-token costs when routing to hosted model APIs

Selection checklist

  1. Decide whether users need a desktop app, a shared web instance, or both, and shortlist accordingly.
  2. Connect the assistant to the model endpoints you already pay for and compare answer quality on real questions.
  3. Verify that single sign-on, roles and audit logs meet your internal policy before a company rollout.
  4. Test document retrieval on a sample of your own files and check that citations point to correct passages.

What belongs in this category: Listed products give end users a ready conversational interface or persistent assistant, self-hostable or local, where the model behind it can be chosen or swapped rather than fixed by a single provider.

AI Assistants & Chat on iThub in figures

iThub lists 41 tools in AI Assistants & Chat. 38 of 41 are open source, 34 of 41 can be self-hosted, 40 of 41 have a free, freemium or open-source pricing model and 22 of 41 offer an API.

Tools listed
41
Open source
38
Self-hostable
34
Free to start
40
With an API
22
Web app
26
Desktop app
30

Recently added: Rakazo, Sugabots, AnythingLLM, GPT4All

Typical use cases

  • Private company chat that keeps prompts off public services
  • Asking questions about internal documentation and policies
  • Comparing answers from several models in one interface
  • Offline assistant use on laptops without network access

Trade-offs to model

  • Desktop apps are quick to try and keep data on one device, while a shared server suits teams but needs administration.
  • Running small local models avoids API costs and data transfer, yet larger hosted models usually answer complex questions better.

Common mistakes

  • Rolling out a shared instance without access controls, so every user can read every uploaded document.
  • Underestimating the memory and GPU needed to serve local models to more than a handful of users.
  • Assuming an assistant that can run tools is safe without reviewing which actions it may take unattended.
Decision resources

Continue the technical evaluation

Use the resources that match your decision stage. Profile data and linked sources were refreshed through November 10, 2026.

Evaluation methodology · Corrections process