This is a module that will interface with [Gemma](https://blog.google/technology/developers/gemma-open-models/)'s [gemma.cpp](https://github.com/google/gemma.cpp). Its still a work in progress.
MediaPipe GenAI Tasks
A high-fidelity agentic terminal assistant for the Flux Era.
Core agent, model providers, and tools for the unofficial Gemma CLI coding agent.
Unofficial local-first coding agent optimized for Gemma models.
Gemma model-family provider for Robota using OpenAI-compatible local endpoints
Node.js package for Gemma 4 — local multimodal inference and agent system
Core types and classes for Google services.
LiteRT-LM package
Node.js Streams, a user-land copy of the stream library from Node.js
On-device Gemma 4 contextual PHI redaction toolkit for HIPAA Safe Harbor de-identification. Runs locally via Ollama (browser WebLLM path wired for when MLC ships a Gemma 4 build). Never on a server.
Reusable OpenAI-compatible Chat Completions transport primitives for Robota providers
Find the fastest coding LLM models in seconds — ping free models from multiple providers, pick the best one for OpenCode, Cursor, or any AI coding assistant.
gemma tokenizer for NodeJS/Browser
Node.js API (Node-API)
Determine if the current node version supports the `--preserve-symlinks` flag.
A light-weight module that brings Fetch API to node.js
Universal AI model proxy — route any coding tool through OpenRouter, Ollama, LMStudio, llama.cpp, or any LLM provider
Load node modules according to tsconfig paths, in run-time or via API.
Picovoice picoLLM Node.js binding
LLM agent loop + remote/WASM/local providers + MCP wrapper
Genkit AI framework plugin for Ollama APIs.
Cross platform child_process#spawn and child_process#spawnSync
Give brains to your small models. MCP server enforcing step-by-step Chain-of-Thought.