DLocker Cache

Collective memory for Ollama —
the more people use it, the faster it gets.

Quick Start

# Linux / macOS
curl -fsSL https://dlocker.app/install.sh | sh

# Windows (PowerShell)
powershell -c "irm dlocker.app/install.ps1 | iex"

# Transparent mode (recommended): intercepts port 11434 automatically.
# Every agent works with zero code changes. Ollama moves to port 11435.
# Standard mode: DLocker runs on port 8000, change base_url manually.

Drop-in Integration

# Before
client = OpenAI(base_url="http://localhost:11434/v1")

# After — that's it.
client = OpenAI(base_url="http://localhost:8000/v1")

# Or use the cloud endpoint:
client = OpenAI(base_url="https://dlocker.app/v1")

Why DLocker

Other cache libraries

pip install + rewrite your LLM calls
Import + modify your code
Does not get faster over time

DLocker

One command to install
Zero code changes
Gets faster over time

Live Stats

—

Cache Hits

—

Cache Misses

—

Hit Rate %

—

Cached Entries

Links

github.com/lu-9999/dlocker-cache (coming soon — repo goes public on launch day)

Support this project → github.com/sponsors/lu-9999