Ollama GUI: Web Interface for chatting with your local LLMs.
Ollama GUI is a web interface for ollama.ai, a tool that enables running Large
Language Models (LLMs) on your local machine.
🛠 Installation
Prerequisites
- Download and install ollama CLI.
- Download and install yarn
and node
ollama pull <model-name>
ollama serve
Getting Started
- Clone the repository and start the development server.
git clone https://github.com/HelgeSverre/ollama-gui.git
cd ollama-gui
yarn install
yarn dev
Or use the hosted web version, by running ollama with the following origin
command (docs)
OLLAMA_ORIGINS=https://ollama-gui.vercel.app ollama serve
Running with Docker
To run Ollama GUI using Docker, follow these steps:
-
Make sure you have Docker (or OrbStack)
installed on your system.
-
Clone the repository:
git clone https://github.com/HelgeSverre/ollama-gui.git
cd ollama-gui
-
Build the Docker image:
docker build -t ollama-gui .
-
Run the Docker container:
docker run -p 8080:8080 ollama-gui
-
Access the application by opening a web browser and navigating to http://localhost:8080
.
Note: Make sure that the Ollama CLI is running on your host machine, as the Docker container for Ollama GUI needs to
communicate with it.
Models
For convenience and copy-pastability
, here is a table of interesting models you might want to try out.
For a complete list of models Ollama supports, go
to ollama.ai/library.
Model |
Parameters |
Size |
Download |
Mixtral-8x7B Large |
7B |
26GB |
ollama pull mixtral |
Phi |
2.7B |
1.6GB |
ollama pull phi |
Solar |
10.7B |
6.1GB |
ollama pull solar |
Dolphin Mixtral |
7B |
4.1GB |
ollama pull dolphin-mixtral |
Mistral |
7B |
4.1GB |
ollama pull mistral |
Mistral (instruct) |
7B |
4.1GB |
ollama pull mistral:7b-instruct |
Llama 2 |
7B |
3.8GB |
ollama pull llama2 |
Code Llama |
7B |
3.8GB |
ollama pull codellama |
Llama 2 Uncensored |
7B |
3.8GB |
ollama pull llama2-uncensored |
Orca Mini |
3B |
1.9GB |
ollama pull orca-mini |
Falcon |
7B |
3.8GB |
ollama pull falcon |
Vicuna |
7B |
3.8GB |
ollama pull vicuna |
Vicuna (16K context) |
7B |
3.8GB |
ollama pull vicuna:7b-16k |
Vicuna (16K context) |
13B |
7.4GB |
ollama pull vicuna:13b-16k |
nexusraven |
13B |
7.4gB |
ollama pull nexusraven |
starcoder |
7B |
4.3GB |
ollama pull starcoder:7b |
wizardlm-uncensored |
13B |
7.4GB |
ollama pull wizardlm-uncensored |
📋 To-Do List
- [x] Properly format newlines in the chat message (PHP-land has
nl2br
basically want the same thing)
- [x] Store chat history using IndexedDB locally
- [x] Cleanup the code, I made a mess of it for the sake of speed and getting something out the door.
- [x] Add markdown parsing lib
- [ ] Allow browsing and installation of available models (library)
- [ ] Ensure mobile responsiveness (non-prioritized use-case atm.)
- [ ] Add file uploads with OCR and stuff.
🛠 Built With
📝 License
Licensed under the MIT License. See the LICENSE.md file for details.