Raffaele Mancini

12 posts

Raffaele Mancini

@RaffoMancio

Trani, Apulia Katılım Ocak 2012

80 Takip Edilen22 Takipçiler

Raffaele Mancini@RaffoMancio·2d

@Yummyfunduk The on-device performance of MedPsy is mind-blowing

English

Alberto Valero@Yummyfunduk·6d

Medical AI must be free, open source and run locally. FULL STOP.

Tether@tether

Tether Unveils Medical AI That Runs on Phones, Outperforms Much Larger SoTA Models, and Can Cut the Cloud Out Entirely Read more: tether.io/news/tether-un…

English

111

Raffaele Mancini retweetledi

Tether@tether·2d

Tether Launches Developer Grants Program to Fund Local-First AI and Payments Infrastructure Read more: tether.io/news/tether-la…

English

455

60.1K

Raffaele Mancini retweetledi

Paolo Ardoino 🤖@paoloardoino·6d

We just released our QVAC MedPsy, Tether AI SoTA medical health AI model, capable of high-performance execution and high-accuracy directly on smartphones, laptops and servers. Highlights: - QVAC MedPsy 4B beats MedGemma 27B - QVAC MedPsy 1.7 beats MedGemma 4B - 3.2x reduction in response tokens, further increasing efficiency - fully open-source - GGUF formats supported - designed to excel on edge devices - 100% user privacy [benchmarked against clinical-style evaluations such as HealthBench Hard, HealthBench, and MedXpertQA]

Tether@tether

8 billion humans deserve an intelligence that doesn't blink when the signal dies. 🧠 Introducing @QVAC Psy, our foundational models built on the mathematical stability of Psychohistory. With QVAC MedPsy, our local-first medical health AI model, we’ve proven that superior methodology beats raw parameter count. Our 1.7B & 4B models are delivering expert-level healthcare reasoning on consumer hardware. The "tiny brain" for the next galaxy is here. Fully open-source. Fully sovereign. Learn more qvac.tether.io/models

English

195

26.1K

Raffaele Mancini@RaffoMancio·7 May

@elchiapp Thanks! This demo is actually based on a custom fork of @qvac where I implemented real time simulated streaming for Parakeet for much better on-device accuracy & multilingual support github.com/Helldez/qvac/c…

English

Marco Chiappetta@elchiapp·6 May

@RaffoMancio @paoloardoino @qvac Great demo! We are shipping a lot of improvements on the speech side in the next couple of versions. Stay tuned!

English

Paolo Ardoino 🤖@paoloardoino·6 May

Your personal composable private AI engine

QVAC@qvac

If your AI needs a wire to work, it isn't yours. QVAC is the local-first engine designed to run anywhere, even where the internet can't reach. Fully autonomous. Fully open source. Fully sovereign. If you can dream it, you can build it. Even in another galaxy. 🛸✨ Start building the future of edge AI: docs.qvac.tether.io 🚀

Italiano

121

20.4K

Raffaele Mancini@RaffoMancio·6 May

@paoloardoino Privacy is a constraint, not a feature. Glad @qvac arrived at the same answer, makes building toward it tractable instead of theoretical

English

Paolo Ardoino 🤖@paoloardoino·6 May

This is the way

Raffaele Mancini@RaffoMancio

@paoloardoino Built on @qvac . Took the "no wires" part literally. Phone in airplane mode running a full STT → LLM → TTS voice loop, multilingual. github.com/Helldez/JarvisQ

English

11.8K

Raffaele Mancini@RaffoMancio·1 May

Tested @qvac on Android: full speech-to-text → LLM → text-to-speech running 100% on-device, no network. Parakeet (STT) + Qwen3 1.7B + Supertonic all via @qvac SDK. Built a small fork open source to make Parakeet run in streaming mode

English

105

Raffaele Mancini retweetledi

Pietro Schirano@skirano·23 Nis

This is the game GPT-5.5 built for me in one shot, featured on the release page. What makes it different from previous AI-generated games is how fun it actually feels to play. The controls are smooth, and every shape is generated from scratch in Three.js.