·  ·    ·        ·     ·    ·  ·
    ·    l l m c h a t    ·
  ·    ·        ·    ·       ·

A terminal chat interface for local LLMs via Ollama.

---

llmchat is a terminal UI for chatting with any model running in Ollama. Conversations persist in a local SQLite database. Responses stream in real time with full markdown rendering. Thinking mode surfaces chain-of-thought output from models that support it.

---

Requirements

• Ollama running locally on port 11434
• At least one model pulled, e.g. ollama pull qwen3:8b

---

Installation

Option 1 - Go install

go install github.com/phlx0/llmchat/cmd/llmchat@latest

Make sure Go's bin directory is on your PATH:

# add to ~/.zshrc or ~/.bashrc
export PATH="$PATH:$(go env GOPATH)/bin"

Option 2 - Pre-built binary (no Go required)

1. Go to the Releases page.

2. Download the archive for your platform:

   OS	Arch	File
   macOS	Apple Silicon	llmchat_darwin_arm64.tar.gz
   macOS	Intel	llmchat_darwin_amd64.tar.gz
   Linux	x86-64	llmchat_linux_amd64.tar.gz
   Linux	ARM64	llmchat_linux_arm64.tar.gz
   Windows	x86-64	llmchat_windows_amd64.zip

3. Extract and move to your PATH:

   tar -xzf llmchat_darwin_arm64.tar.gz
   sudo mv llmchat /usr/local/bin/
   llmchat --version

Option 3 - Build from source

git clone https://github.com/phlx0/llmchat
cd llmchat
go install ./cmd/llmchat

Requires Go 1.26+.

---

Usage

llmchat

Conversations save to ~/.local/share/llmchat/llmchat.db.

---

Keys

Chat

Key	Action
Enter	Send message
Alt+Enter	Newline in input
Tab	Open conversation list
Ctrl+P	Pick model
Ctrl+T	Toggle thinking mode
Ctrl+Y	Toggle thinking visibility
Ctrl+C	Quit

Conversation list

Key	Action
Enter	Open conversation
n	New conversation
d	Delete conversation
/	Filter
Esc / Tab	Back to chat

---

Config

~/.config/llmchat/config.json

{
  "model": "qwen3:8b"
}

Defaults to qwen3:8b if absent. Ctrl+P writes the selected model automatically.

---

Star History

---

MIT License · made by phlx0