volcengine/OpenViking
OpenViking is an open-source context database designed specifically for AI Agents(such as openclaw). OpenViking unifies the management of context (memory, resources, and skills) that Agents need through a file system paradigm, enabling hierarchical context delivery and self-evolving.
Implements a three-tier hierarchical storage system (L0/L1/L2) with on-demand loading to reduce token consumption, and combines directory-based retrieval with semantic search for precise context acquisition. Built with Python bindings atop Go/C++ core components (AGFS), it integrates with multiple VLM providers (Volcengine, OpenAI, LiteLLM) and includes automatic session management for extracting and compressing long-term agent memory. Provides visualization of retrieval trajectories to make the context retrieval pipeline observable and debuggable.
7,606 stars and 81,507 monthly downloads. Used by 1 other package. Actively maintained with 371 commits in the last 30 days. Available on PyPI.
Stars
7,606
Forks
541
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 13, 2026
Monthly downloads
81,507
Commits (30d)
371
Dependencies
38
Reverse dependents
1
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/volcengine/OpenViking"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Featured in
Related tools
MemoriLabs/Memori
SQL Native Memory Layer for LLMs, AI Agents & Multi-Agent Systems
zjunlp/LightMem
[ICLR 2026] LightMem: Lightweight and Efficient Memory-Augmented Generation
mem0ai/mem0
Universal memory layer for AI Agents
memodb-io/memobase
User Profile-Based Long-Term Memory for AI Chatbot Applications.
agentscope-ai/ReMe
ReMe: Memory Management Kit for Agents - Remember Me, Refine Me.