AI/ML●●●Banger
ChonkLM – Tiny language models running offline in the browser
Runs GGUF models in the browser via custom WGSL shaders when cloud APIs ignore tiny models.
Zero to OneWizardryNiche Gem
bilalba
601mo ago
Back to browse
LocalVQE: Tiny ~1M param audio model that cancels echo and noise
by richiejp·May 1, 2026·1 point·0 comments
AI Analysis
●●●BangerWizardryBig BrainSolve My Problem
1M parameters doing what Krisp needs millions for, running entirely on CPU.
Strengths
- •Single-pass architecture handles echo cancellation, noise suppression, and dereverberization together.
- •Tiny model size (~1M params) enables real-time CPU inference without GPU dependency.
- •Bundled test clips let you verify far-end echo cancellation before committing to integration.
Weaknesses
- •Hugging Face Spaces demo limits real-world latency testing for interactive use cases.
- •No mentioned integration examples for popular frameworks like WebRTC or Unity.
Category
Target Audience
Voice app developers, remote work tool builders
Similar To
Krisp · NVIDIA Maxine · RNNoise
Similar Projects
Developer Tools●●Solid
Pending – a tiny pure-Go in-memory deferred task scheduler
Go scheduler with ID-based debouncing, but crowded space already has cron and job queues.
Niche GemCozy
kahoonster
213mo ago
Productivity●●Solid
Tanin – TUI Noise Generator
Ratatui TUI wraps ambient sounds with yt-dlp import, but Blanket GUI already exists.
CozyNiche GemSlick
anongazelle
824mo ago
AI/ML●Mid
Vynix – Mobile app with 100 AI models for image/video/audio generation
100 models in one app—but mobile AI aggregators already exist; no novel moat here.
Crowd Pleaser
cosmicmeta
123mo ago
Productivity●Mid
Transcribe audio from apps, files and your device's microphone
Free offline Whisper on Android, but 88–93% accuracy lags Google Recorder's 95–96% by design constraints.
Ship It
dstudios
103mo ago
AI/ML●●Solid
I made a private/local audio transcriber
Local Whisper in browser when Whisper.cpp and web-whisper already exist.
Solve My ProblemCozy
fzsheikdev
201mo ago