Works with LlamaIndex

LlamaIndex, running on your Mac

ToolPiper is a LlamaIndex LLM. Index, retrieve and query without a document leaving the machine.

Download ToolPiper Read the API docs

Free, Apple Silicon, signed. The local model runner has no caps and needs no account.

Setup

  1. Install ToolPiper and launch it once. The local API listens on port 9998.
  2. Install the package: `pip install llama-index-llms-toolpiper`.
  3. Use it anywhere LlamaIndex takes an `llm`.
from llama_index.llms.toolpiper import ToolPiper

llm = ToolPiper(model="gemma-4-12b-it-Q4_K_M")
print(llm.complete("Summarise this in one sentence."))

Prompts, responses and files stay on the machine. There is no key to rotate and no per-token cost. LlamaIndex talks to ToolPiper over its local OpenAI-compatible API on port 9998.