openviking
An Agent-native context database
Decision gist · record as of 2026-08-14
Yes, with conditions. OpenViking is actively maintained with recent releases and solves a real problem for AI agent developers—observable, token-efficient context retrieval. However, the AGPL-3.0 license requires copyleft compliance in derivative works, and the 58 runtime dependencies create medium install friction. Install if you are building agents that need persistent, debuggable memory and can accept the license terms; avoid if you need a permissive-licensed context store or want minimal dependencies.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.10 or higher.
- The `init` wizard requires interactive setup of an LLM provider (Volcengine, OpenAI, Ollama, etc.) and writes configuration to ~/.openviking/ov.conf before the server can start.
- Medium install friction due to 58 runtime dependencies including openai, fastapi, scrapy, and pdfplumber.
License · maintenance · safety
AGPL-3.0 (agpl) — Licensed under AGPL-3.0, which requires that any modifications or derivative works distributed must also be released under AGPL-3.0. This is a copyleft license; using this package in a proprietary application requires careful consideration of your distribution model.
last release 2026-08-06 (8 days) · last repo commit 2026-08-14 · 28,446 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 400,754 downloads/mo, #6,935 on PyPI
Alternatives
Verify before relying
pip install openviking
openviking-server init
openviking-server
# In another terminal:
ov add-resource https://github.com/example/repo
ov find "search query"- Whether the 58 runtime dependencies can be selectively installed or if all are required for basic operation
- Performance characteristics and absolute throughput of retrieval at scale
- Whether AGPL-3.0 copyleft applies to agents using OpenViking as a service vs. bundled deployment
What it is and what it does
OpenViking is an open-source context database designed for AI agents that replaces traditional vector stores with a filesystem-like interface. Instead of querying a black-box retrieval system, agents browse context using familiar commands (`ls`, `tree`, `find`, `grep`) on a `viking://` protocol namespace. Content is processed into three layers on write—L0 (abstract, ~100 tokens), L1 (overview, ~2k tokens), and L2 (full details)—and loaded on demand to reduce token spend and latency.
The package includes a server component, a CLI client, and integrations for popular AI agents. It supports ingesting resources from URLs, GitHub repos, and local files, then automatically extracts and commits session memory after agent interactions. Retrieval is observable: each query preserves its directory-browsing trajectory so you can debug why a result was retrieved. The fact sheet shows benchmarks reporting 80–83% accuracy on user memory tasks with 34.3–91.0% reduction in input tokens and 58.45–66.10% latency improvement.
Use it for
- Build AI agents with persistent, searchable memory that survives across sessions and reduces token consumption on long conversations.
- Ingest and organize technical documentation, code repositories, and web content as agent-accessible context with semantic search.
- Debug agent retrieval behavior by inspecting the directory-browsing trajectory that produced each result.
- Integrate OpenViking recall into existing agent frameworks via provided plugins and integrations.
- Extract and store agent experience and user preferences as long-term memory after multi-turn sessions.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, with conditions.
OpenViking is actively maintained with recent releases and solves a real problem for AI agent developers—observable, token-efficient context retrieval. However, the AGPL-3.0 license requires copyleft compliance in derivative works, and the 58 runtime dependencies create medium install friction. Install if you are building agents that need persistent, debuggable memory and can accept the license terms; avoid if you need a permissive-licensed context store or want minimal dependencies.
Install
openviking on PyPI
Before you install
Medium install friction due to 58 runtime dependencies including openai, fastapi, scrapy, and pdfplumber. Active maintenance with recent releases (8 days old) and strong community signal. Requires Python 3.10 or higher and includes a CLI setup wizard for provider configuration.
Requires Python 3.10 or higher. The `init` wizard requires interactive setup of an LLM provider (Volcengine, OpenAI, Ollama, etc.) and writes configuration to ~/.openviking/ov.conf before the server can start.
License in practice
Licensed under AGPL-3.0, which requires that any modifications or derivative works distributed must also be released under AGPL-3.0. This is a copyleft license; using this package in a proprietary application requires careful consideration of your distribution model.
Quickstart
pip install openviking
openviking-server init
openviking-server
# In another terminal:
ov add-resource https://github.com/example/repo
ov find "search query"Verify before relying
- Whether the 58 runtime dependencies can be selectively installed or if all are required for basic operation
- Performance characteristics and absolute throughput of retrieval at scale
- Whether AGPL-3.0 copyleft applies to agents using OpenViking as a service vs. bundled deployment
Package facts
| License | AGPL-3.0 agpl |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 58 packagesopenviking-sdkpydantictyping-extensionspyyamlhttpxpdfplumberscrapytrafilaturafeedparserdefusedxmlopenairequestscharset-normalizerpython-docxolefilexlrdpython-pptxopenpyxlebooklibjson-repairapschedulervolcenginevolcengine-python-sdkfastapiuvicornxxhashjinja2tabulateurllib3protobuf |
| Maintenance | Actively maintained 8 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 400,754 / month, #6,935 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 3 - AlphaIntended Audience :: DevelopersProgramming Language :: Python :: 3Programming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Topic :: Scientific/Engineering :: Artificial Intelligence |
Evidence: openviking-0.4.13-cp310-abi3-macosx_14_0_arm64.whl; openviking-0.4.13-cp310-abi3-macosx_15_0_x86_64.whl; openviking-0.4.13-cp310-abi3-manylinux_2_31_aarch64.whl; openviking-0.4.13-cp310-abi3-manylinux_2_31_x86_64.whl; openviking-0.4.13-cp310-abi3-win_amd64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “context database for AI agents”
- openvikingOpenViking is a context database for AI agents that organizes…
- memoriMemori is a Python SDK that automatically captures and recalls…
- cogneeCognee builds a self-hosted knowledge graph from ingested data,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also memsearch · omnigent · mem0ai · stashai · openhands-agent-server · ouroboros-ai · agent-framework-azure-ai · openhands-workspace · langmem · agent-framework-mem0