Practical overview
When it fits
Use Wikichat when you want to inspect or self-host a chatbot that grounds language-model responses in Wikipedia or a document collection you index.
Who it is for
- Developers building or adapting corpus-grounded question-answering applications
- Researchers exploring retrieval-augmented chat and factuality
- Teams that want to connect a chatbot to a custom document index
Documented uses
- Ask questions with answers supported by retrieved Wikipedia material
- Configure a language model and an information-retrieval source for a chatbot
- Build or host a retrieval index for custom documents
- Explore a research implementation of retrieval-augmented conversation
Requirements & limits
The repository identifies the project as open source under the Apache-2.0 license and links to an online demo.
- The chatbot requires both a configured language model and a retrieval source.
- The README lists more than 100 supported language models, including options from OpenAI, Azure, Anthropic, Mistral, HuggingFace, Together.ai, and Groq.
- Retrieval options include a rate-limited Wikipedia API, a hosted Wikipedia index, and a new index built from custom documents.
- Basic use with LLM APIs and the Wikipedia search API has minimal hardware requirements; local models and creating new indexes have GPU requirements.
- The repository provides optional multi-user front-end and back-end code and instructions for Azure Cosmos DB conversation storage.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Prepare a self-hosted Wikichat setup
Set up the documented project environment and identify the model and retrieval components to configure before running Wikichat.
Before you start
- Git and Pixi installed
- Python 3.11 is recommended; the project was tested on Ubuntu 20.04 LTS
- A supported language model choice
- A retrieval source: the Wikipedia search API, a hosted index, or a custom index
Steps
Clone the WikiChat repository and change into its directory.
Create and activate the environment specified by the repository using
pixi shell.Choose a supported language model and configure it as described in the repository.
Select a retrieval source and configure its endpoint according to the repository instructions.
Expected outcome
The recommended environment is prepared and the model and retrieval choices needed for a configured WikiChat run are identified.
Guide limitations
- This guide covers setup and configuration choices, not a verified successful run.
- Windows WSL, macOS, and Python versions other than 3.11 may require troubleshooting.
- A local language model requires a GPU; creating a new index requires a GPU, and the default embedding model needs at least 13 GB of GPU memory.
Compare options for Retrieval-augmented applications
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| Wikichat | Ask questions with answers supported by retrieved Wikipedia material | Self-hosted, Web | The repository identifies the project as open source under the Apache-2.0 license and links to an online demo. |
| Rag Powered AI Assistant Transforming Data Retrieval And Analysis Across The Web And Pdfs | Retrieve relevant passages from web and PDF content to answer a question. | Web | |
| LangChain in Supabase Edge Functions | Initialize and start a Supabase project locally. | Cloud, Server |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 8 October 2026
- Original website title
- stanford-oval/WikiChat · Official homepage title · retained in its source language