RepoRadar
Leave a Tip
Item detail

tetherto/qvac

QVAC is an open-source, cross-platform, peer-to-peer AI ecosystem from Tether that runs LLMs, speech, and RAG locally across Linux / macOS / Windows / Android / iOS — or delegates inference to peers using built-in P2P capabilities (BitTorrent-style for AI). 255 stars, Apache-2.0, JavaScript SDK, OpenAI-compatible HTTP API, runs on Node.js, Bare runtime, and Expo.

Open source →Back to radar
Score7.7
Popularity68.0
Riskconditional
TierSilver
Score breakdown
Usefulness7.0
Novelty9.0
Momentum6.0
Maturity7.0
Open-source/build8.4
Evidence7.2
Workflow potential8.1
Setup ease8.8

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for mobile and cross-platform developers who want to run AI inference locally on iOS / Android / desktop with a single JavaScript SDK, or build P2P AI apps that delegate inference across devices without a central API.

Who should use it

mobile developers who want to run AI inference on iOS / Android with a single JavaScript SDKcross-platform teams building React Native / Expo apps that need on-device LLM / speech / RAGdevelopers building P2P AI apps that delegate inference across devices without a central serverteams that want an OpenAI-compatible API endpoint running entirely on-deviceresearchers experimenting with decentralized inference topologies

Who should skip it

Skip if the source link, docs, or setup requirements do not match your workflow.

Risk explanation

Maintained by Tether (a crypto company) — vet roadmap and licensing stability before depending on it for production; P2P delegation model is novel — security, performance, and reliability characteristics are not yet battle-tested; Mobile (iOS / Android) on-device inference has tight memory / compute constraints — benchmark against llama.cpp / Ollama before assuming parity.

Evidence links

Closest alternatives / related signals

local-aip2pinferencecross-platformiosandroidreact-nativeexpo