DEDUPEDWho it is forA team that wants agent memory and knowledge on its own infrastructure.
The problemHosted memory services keep your project context on someone else's servers.
The outcomeAn OpenViking server with ~/.openviking/ov.conf set for your chosen provider, including local Ollama.
What you need first
- uv and Python for a self-hosted server, or the hosted OpenViking Service
- A model provider for the server (Volcengine, OpenAI, Codex OAuth, Kimi, GLM or local Ollama)
- A user key for the memory plugin (a root key cannot read or write memories)
Steps
- Run
uv tool install openviking --upgrade && openviking-server init - Choose a model provider in the wizard: Volcengine, OpenAI, Codex OAuth, Kimi, GLM or local Ollama
- Start the server and check it with
ov status - Decide whether other machines need to reach it and whether auth is on
Prompt for your agent
Self-host volcengine/OpenViking on <machine>: install it with uv, run openviking-server init with <provider>, start it and confirm ov status works. Tell me the address and whether auth is on.
Caveat.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for volcengine/OpenViking and "Self-host an OpenViking context server with your own models"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer