NobodyWho
Open-source, free inference engine for running LLMs fully on-device — no cloud, no API keys — across mobile, desktop, and game engines.
🔗 Visit NobodyWho
Description
Running an AI model usually means sending your data to someone else's server and paying per request. NobodyWho flips that: the model runs directly on the user's own phone, computer, or even smartwatch, with nothing leaving the device.
NobodyWho is an open-source inference engine, built on llama.cpp, that runs large language models locally and efficiently on any device without cloud services, API keys, or internet connectivity. It supports text, vision, and speech models with GPU-accelerated inference (Metal, CUDA, Vulkan), and ships native SDKs for Swift, Kotlin, Flutter, React Native, Python, and Godot — making it usable across iOS, Android, Windows, Linux, and even game engines and smartwatches. It's free and open source under the EUPL 1.2 license, with around 1,100 GitHub stars.
💬 Our review
The short version: NobodyWho is a solid, genuinely free option for developers who want on-device AI without the cost or privacy trade-offs of a cloud API — the multi-platform SDK support (including Godot) is its most distinctive feature.
Against Ollama, LM Studio, or LocalAI, all more established local-inference tools, NobodyWho's differentiator is breadth of platform support beyond desktop — native SDKs for mobile (Swift, Kotlin, Flutter, React Native) and game engines (Godot) are less common in this space, where most local-inference tools focus on desktop use. That makes it a more natural fit for building an AI feature into a mobile app or game than adapting a desktop-first tool. The trade-off is community size: at roughly 1,100 GitHub stars, it's meaningfully smaller than Ollama, which means fewer community models, tutorials, and third-party integrations to lean on. It also requires GGUF-format models, so you're constrained to what's available in that ecosystem. Best fit: developers building mobile apps or games that need offline, privacy-preserving AI without a subscription or API key. Weaker fit: teams wanting the largest community and most plug-and-play model library — Ollama still has the edge there.
💰 Pricing
📊 Global score
🤖 AI-enriched data
100% gratuit, open source sous licence EUPL 1.2, aucune clé API ni coût d'abonnement.
Pros
Fonctionne entièrement hors-ligne, sans clé API
SDKs natifs multi-plateformes (iOS, Android, Windows, Linux, montres, Godot)
Inférence accélérée GPU (Metal, CUDA, Vulkan)
Confidentialité : toutes les données restent sur l'appareil
Gratuit et open source (EUPL 1.2)
Cons
Communauté plus petite que Ollama (~1 100 stars)
Nécessite des modèles au format GGUF
Performance dépendante du matériel de l'appareil
Projet encore jeune, développement actif