Today
Intel
Market
Earn
Settings
Account
Theme Selection
Light
Dark
Language
English
简体中文
繁體中文
Tiếng Việt
한국어
Followin APP
Mine Web3 Possibilities
App Store
Google Play
Log in
NodeShift
53,904 Twitter followers
Follow
A decentralized, no-code AI cloud platform that enables one-click deployment of AI agents and LLMs in a fully private and sovereign environment.
Posts
NodeShift
09-18
Thread
#Thread#
Introducing Tongyi DeepResearch (30B-A3B) – Alibaba’s Breakthrough in Agentic AI
MODE
4.59%
NodeShift
09-10
Thread
#Thread#
Baidu, Inc. just dropped another open-weight beast — ERNIE-4.5-21B-A3B-Thinking — a 21B parameter Mixture-of-Experts (MoE) model with 3B active experts/token, optimized for reasoning, coding, long-context, and function-calling.
TUT
1.35%
NodeShift
05-20
Thread
#Thread#
What if AI could write advance algorithms, plus test and continuously evolve them, on its own? That's exactly what AlphaEvolve by @GoogleDeepMind brings to the table.
TST
1.94%
NodeShift
05-19
Thread
#Thread#
Say hello to SmolDocling-256M-preview — a compact but highly capable document understanding model that turns scanned PDFs, academic papers, invoices, and forms into clean, structured, readable text using rich DocTags. #smoldocling #AIModels
TUT
1.35%
NodeShift
05-17
Thread
#Thread#
Alibaba's Wan2.1 Just Leveled Up! Wan2.1 VACE is here, the latest open-source model in the popular Wan2.1 series, and this model will let you turn your text into cinematic videos, edit any scene, or even generate realistic sound from video - all using AI. Read more⬇️
NodeShift
05-16
Thread
#Thread#
Qwen3-32B is a powerful and thoughtful model designed to move effortlessly between deep reasoning and light conversation. Whether it's solving math step-by-step, writing stories, or switching between 100+ languages — it does it all with clarity and purpose. #qwen32b #qwen3
NodeShift
05-15
Thread
#Thread#
Apple just dropped a lightening-fast Vision-Language Model and it's a game changer. Meet FastVLM: - 85x faster Time-to-First-Token than LLaVA - 3.4x smaller vision encoder - Real-time performance on mobile - Built for high-res image input and low-latency inference Get started ⬇️
Loading..