Vessel

A modern, feature-rich web interface for local LLMs

Features • Quick Start • Documentation • Contributing

---

Why Vessel

Vessel is intentionally focused on:

• A clean, local-first UI for local LLMs
• Multiple backends: Ollama, llama.cpp, LM Studio
• Minimal configuration
• Low visual and cognitive overhead
• Doing a small set of things well

If you want a universal, highly configurable platform → open-webui is a great choice. If you want a small, focused UI for local LLM usage → Vessel is built for that.

---

Features

Chat

• Real-time streaming responses with token metrics
• Message branching — edit any message to create alternative conversation paths
• Markdown rendering with syntax highlighting
• Thinking mode — native support for reasoning models (DeepSeek-R1, etc.)
• Dark/Light themes

Projects & Organization

• Projects — group related conversations together
• Pin and archive conversations
• Smart title generation from conversation content
• Global search — semantic, title, and content search across all chats

Knowledge Base (RAG)

• Upload documents (text, markdown, PDF) to build a knowledge base
• Semantic search using embeddings for context-aware retrieval
• Project-specific or global knowledge bases
• Automatic context injection into conversations

Tools

• 5 built-in tools: web search, URL fetching, calculator, location, time
• Custom tools: Create your own in JavaScript, Python, or HTTP
• Agentic tool calling with chain-of-thought reasoning
• Test tools before saving with the built-in testing panel

LLM Backends

• Ollama — Full model management, pull/delete/create custom models
• llama.cpp — High-performance inference with GGUF models
• LM Studio — Desktop app integration
• Switch backends without restart, auto-detection of available backends

Models (Ollama)

• Browse and pull models from ollama.com
• Create custom models with embedded system prompts
• Per-model parameters — customize temperature, context size, top_k/top_p
• Track model updates and capability detection (vision, tools, code)

Prompts

• Save and organize system prompts
• Assign default prompts to specific models
• Capability-based auto-selection (vision, code, tools, thinking)

📖 Full documentation on the Wiki →

---

Screenshots

Clean chat interface	Syntax-highlighted code
Integrated web search	Model browser

---

Quick Start

Prerequisites

• Docker and Docker Compose
• An LLM backend (at least one):
  • Ollama (recommended)
  • llama.cpp
  • LM Studio

Configure Ollama

Ollama must listen on all interfaces for Docker to connect:

# Option A: systemd (Linux)
sudo systemctl edit ollama
# Add: Environment="OLLAMA_HOST=0.0.0.0"
sudo systemctl restart ollama

# Option B: Manual
OLLAMA_HOST=0.0.0.0 ollama serve

Install

# One-line install
curl -fsSL https://somegit.dev/vikingowl/vessel/raw/main/install.sh | bash

# Or clone and run
git clone https://github.com/VikingOwl91/vessel.git
cd vessel
./install.sh

Open http://localhost:7842 in your browser.

Update / Uninstall

./install.sh --update     # Update to latest
./install.sh --uninstall  # Remove

📖 Detailed installation guide →

---

Documentation

Full documentation is available on the GitHub Wiki:

Guide	Description
Getting Started	Installation and configuration
LLM Backends	Configure Ollama, llama.cpp, or LM Studio
Projects	Organize conversations into projects
Knowledge Base	RAG with document upload and semantic search
Search	Semantic and content search across chats
Custom Tools	Create JavaScript, Python, or HTTP tools
System Prompts	Manage prompts with model defaults
Custom Models	Create models with embedded prompts
Built-in Tools	Reference for web search, calculator, etc.
API Reference	Backend endpoints
Development	Contributing and architecture
Troubleshooting	Common issues and solutions

---

Roadmap

Vessel prioritizes usability and simplicity over feature breadth.

Completed:

• Multi-backend support (Ollama, llama.cpp, LM Studio)
• Model browser with filtering and update detection
• Custom tools (JavaScript, Python, HTTP)
• System prompt library with model-specific defaults
• Custom model creation with embedded prompts
• Projects for conversation organization
• Knowledge base with RAG (semantic retrieval)
• Global search (semantic, title, content)
• Thinking mode for reasoning models
• Message branching and conversation trees

Planned:

• Keyboard-first workflows
• UX polish and stability improvements
• Optional voice input/output

Non-Goals:

• Multi-user systems
• Cloud sync
• Plugin ecosystems
• Cloud/API-based LLM providers (OpenAI, Anthropic, etc.)

Do one thing well. Keep the UI out of the way.

---

Contributing

Contributions are welcome!

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes
4. Push and open a Pull Request

📖 Development guide →

Issues: github.com/VikingOwl91/vessel/issues

---

License

GPL-3.0 — See LICENSE for details.

Made with Svelte • Supports Ollama, llama.cpp, and LM Studio