Practical overview
When it fits
Use it when you want a self-managed application example for uploading documents, embedding them through the Cosmos API and querying indexed content in chat.
Who it is for
- Researchers looking to retrieve information from collections of textual files
- Content managers handling document collections
- Developers prototyping a Cosmos API-based document indexing and chat application
Documented uses
- Create, rename and manage named file indexes.
- Upload or remove files from an index.
- Embed indexed content to enable RAG retrieval.
- Ask questions about an index through a natural-language chat interface.
Requirements & limits
The README requires a Cosmos API key and instructs users to sign up or log in to the Cosmos platform to generate one. The repository is public and MIT-licensed.
- RAG querying requires embedding the index first.
- The documented development setup requires Node.js 18+, pnpm, Python 3.9+, Poetry and a Cosmos API key.
- Index deletion is delayed by a two-hour countdown, during which restoration is possible.
- The frontend uses Next.js 15, Tailwind CSS, shadcn/ui and TanStack Query; the backend uses FastAPI and the Delos Intelligence Cosmos client.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up the local development application
Prepare the documented frontend and backend development environment and configure its Cosmos API key.
Before you start
- Node.js 18 or later
- pnpm
- Python 3.9 or later
- Poetry
- A Cosmos API key generated from the platform dashboard
Steps
Clone the repository and enter its directory.
Run pnpm install from the frontend directory.
Run poetry install from the backend directory.
Create a root .env file based on .env.example and add COSMOS_API_KEY=your_api_key_here.
Expected outcome
Frontend and backend dependencies are installed and the application has the API key setting described in the README.
Guide limitations
- These steps cover setup and configuration; they do not establish that the application was run or tested.
- The README documents starting the frontend and backend as separate development servers.
Create and query an embedded document index
Use the application workflow to create an index, prepare its contents for RAG, and chat with those files.
Before you start
- An available Cosmos API key configured for the application
- The application accessed at its documented local frontend address or another configured address
Steps
Create a named index and upload files.
Select the index and embed its contents; the README identifies embedding as required for RAG operations.
Start chatting with the selected index.
Expected outcome
The documented workflow lets users query indexed file contents through the chat interface, with responses described as sourced using RAG.
Guide limitations
- The README does not specify supported file formats.
- Deleting an index is scheduled after a two-hour countdown; restoration is available during that countdown.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 16 October 2026
- Original website title
- Delos-Intelligence/cosmos-api-index · Official homepage title · retained in its source language