nobodywho-ooo/nobodywho

Health v2 · maintenance only

Adoption not part of the score

1086 stars · 77 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

NobodyWho is an open-source (EUPL-1.2) on-device LLM inference engine written in Rust, built on llama.cpp, with SDKs for Kotlin, Swift, Python, Flutter, React Native, and Godot. It supports any GGUF chat model, GPU acceleration via Vulkan/Metal, tool calling, multimodal input, plus Whisper speech-to-text and local text-to-speech.

Use cases

• run LLMs locally on mobile apps without a server
• add offline chatbot AI to a Godot game
• embed on-device AI in a Flutter or React Native app
• transcribe audio to text locally with Whisper
• synthesize speech from text on-device
• run GGUF models from Hugging Face with GPU acceleration
• add type-safe tool calling to a local LLM

When to choose

• you need offline, privacy-preserving LLM inference on mobile or desktop
• you want to avoid API costs and servers by running models on-device
• you're building with Godot, Flutter, React Native, Kotlin, Swift, or Python
• you need speech-to-text or text-to-speech bundled with LLM inference

When to avoid

• you need large models that exceed on-device memory and require cloud GPUs
• you need a managed inference service with autoscaling
• your stack isn't among the supported SDKs

Facets

library · maturity active

llm-inference speech-recognition tts sdk machine-learning large-language-models artificial-intelligence mobile-development speech-processing cross-platform cross-platform python rust jvm game-engine on-device-ai gguf llama.cpp godot-plugin flutter react-native offline-inference tool-calling multimodal game-development android ios swift desktop

Member repositories

For agents

markdown · JSON · MCP: product_card(name="nobodywho-ooo/nobodywho")