Deploy Now

SaaS Alternative

Pinecone Weaviate Cloud Zep Cloud

Categories

Developer Tools AI Database Machine Learning Knowledge Base RAG Agent Memory

Stars

36,354

Forks

2,778

Watchers

105

Developer links

Last Updated: 09-10-2026
Share on X

OpenViking

OpenViking is a self-hosted context database that gives AI agents persistent, structured memory by organizing knowledge, skills, and session history into a hierarchical virtual filesystem accessible through the viking:// URI protocol. Instead of dumping everything into a flat vector store and hoping semantic search finds the right chunks, agents navigate their context with familiar commands like ls, tree, and find, locating exactly the information they need through deterministic paths combined with semantic search. Every resource is automatically processed into three layers: a 100-token L0 abstract for quick filtering, a 2,000-token L1 overview for content navigation, and the full L2 detail loaded only when confirmed necessary. This tiered approach cuts token consumption by 83 to 96 percent compared to conventional RAG while improving task completion rates by 15 to 49 percent on benchmark tests. The built-in memory self-iteration loop automatically analyzes task execution and user feedback, updating agent memory directories so the system continuously learns and improves. You can connect to any LLM provider, including Ollama for fully local inference, OpenAI, or compatible gateways. The Web Studio UI at the /studio endpoint provides visual browsing of the entire context filesystem, and the REST API on port 1933 supports programmatic access. Deploy via Docker, Kubernetes with the included Helm chart, or as a standalone service. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

OpenViking
OpenViking
OpenViking
OpenViking
OpenViking
OpenViking

Benefits

  • Slashes Token Costs Up to 96%
  • The three-tier L0/L1/L2 loading architecture delivers only the context layer each query needs, reducing token consumption by 83 to 96 percent compared to flat vector retrieval systems that load entire document chunks.
  • Filesystem Makes Context Traceable
  • Every retrieval follows a visible path through the viking:// directory tree, producing a full trajectory log so you can debug exactly which context was loaded, when, and why.
  • Self-Evolving Agent Memory
  • The built-in iteration loop analyzes completed tasks and user feedback, then automatically updates memory directories so agents grow smarter with each interaction instead of starting fresh every session.
  • Full LLM Provider Flexibility
  • Connect to Ollama for fully local inference, OpenAI, Volcengine Ark, Kimi, GLM, or any OpenAI-compatible endpoint without changing the context layer. Swap providers with a single config change.

Features

  • Viking Protocol Filesystem
  • Organize agent memory, knowledge RAG, and skills under deterministic viking:// URIs with ls, tree, find, and recursive directory retrieval for precise context access.
  • Three-Tier Context Loading
  • L0 abstracts of roughly 100 tokens for filtering, L1 overviews of roughly 2,000 tokens for navigation, and full L2 content loaded on demand to minimize token spend.
  • Web Studio Interface
  • Visual browser at /studio for exploring the context filesystem, monitoring retrieval trajectories, and managing agent sessions through an interactive web dashboard.
  • Docker and Helm Deployment
  • Pre-built GHCR images run the HTTP service on port 1933 with a single docker run command. Helm chart included for Kubernetes clusters.
  • Multi-Provider LLM Support
  • Compatible with Ollama, OpenAI, Volcengine Ark, Kimi, GLM, and any OpenAI-compatible endpoint. Automatic hardware detection for local model selection.