ollaya-dev/ollaya
Local Jev-compatible API server that pulls and serves open decision models for typed, calibrated answers in milliseconds.
Tested on Linux
Use it
Linux x86_64
You need first
- Runtimes:
docker
Download install script from ollaya.dev
curl -fsSL https://ollaya.dev/install.sh -o /tmp/install.shRun install script to install ollaya to ~/.local
bash /tmp/install.sh
Check that it works
ollaya --version 2>&1; ollaya -v 2>&1
ollaya help 2>&1The output should contain “Run open decision models locally”.
First thing to try
ollaya servecurl -s http://127.0.0.1:11435/api/tags 2>&1; echo "---"; curl -s http://127.0.0.1:11435/api/ps 2>&1; echo "---"; curl -s http://127.0.0.1:11435/v1/models 2>&1Add it to your agent
claude mcp add ollaya -- ollaya mcpJSON for Claude Desktop, Cursor and other clients
{
"mcpServers": {
"ollaya": {
"command": "ollaya",
"args": [
"mcp"
]
}
}
}What it does
- Expose models to AI agents via MCP serverTested
- Pull and serve open decision models locally via CLI and daemonNot shown working
- Serve TypeSafe Jev-compatible API at /v1/systemoneNot tested
- Return calibrated probabilities for typed decisions without generating textNot tested
How far pickfit got
Linux x86_64
- Installs
- Connects
- 1 of 4 features work
- Works offline
Windows: the project says it runs there; pickfit hasn’t tried it.
Good for
- serve open decision models locally behind a Jev-compatible API
An alternative to TypeSafe Jev and Ollama.
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
An MCP client called one of its tools
- Replayed from scratch
- Works offline
- Checked on 2 Nov 2026
- Version 0.18.0
An MCP client handshake over stdio listed 4 tools, and a call to list_models returned a result.
Differences the test machine hit
- README: curl -fsSL https://ollaya.dev/install.sh | sh
In the test: Download script first to /tmp/install.sh, read it, then run bash /tmp/install.sh; install auto-falls back to ~/.local without root