AI/ML●●●Banger
Apple's Sharp Running in the Browser via ONNX Runtime Web
Runs Apple's 2.4GB SHARP model entirely in-browser using ONNX Runtime Web.
WizardryBig BrainNiche Gem
bring-shrubbery
185461mo ago
Back to browse
A magic eraser that runs a 208MB ONNX model in the browser
by shadoxise·Apr 15, 2026·2 points·0 comments
AI Analysis
●●SolidSlickSolve My Problem
Runs 208MB LaMa model locally via WebGPU, keeping images off servers.
Strengths
- •WebGPU inference runs the 208MB LaMa model client-side with no server costs.
- •Privacy-first architecture ensures user images never leave the browser environment.
- •Adjustable brush controls allow precise masking before AI reconstruction begins.
Weaknesses
- •Initial 200MB model download creates significant latency on slow connections.
- •CPU fallback takes 10-30 seconds, frustrating users on older hardware.
Category
Target Audience
Designers, privacy-conscious users, quick-edit needs
Similar To
Cleanup.pictures · Magic Studio · PhotoRoom
Similar Projects
AI/ML●●Solid
I taught LLMs to play Magic: The Gathering against each other
LLMs play full-rules Magic: The Gathering with no simplified rulesets or shortcuts.
Big BrainRabbit Hole
GregorStocks
117833mo ago
AI/ML●●●Banger
Real-time local TTS (31M params, 5.6x CPU, voice cloning, ONNX)
5.6x realtime on CPU with voice cloning beats most local TTS options.
WizardryDark Horse
ZDisket
442mo ago
AI/ML●●●Banger
TiGrIS, a tiling compiler that fits ML models onto embedded devices
Fits 4.5MB models into 256KB SRAM via tiling with zero dynamic memory allocation.
Big BrainNiche Gem
asteinh
2001mo ago
AI/ML●●●Banger
Three new Kitten TTS models – smallest less than 25MB
SOTA expressivity at 14M parameters beats cloud models for on-device TTS.
WizardryNiche GemZero to One
rohan_joshi
5611812mo ago
Developer Tools●●Solid
Weight-inspect, examine popular ML models formats without the weights
Deterministic fingerprinting for model structure without loading weights.
Niche GemSolve My Problem
sakuraiben
103mo ago