I Tested the 1-Bit LLM That Fits in 1 GB — It Shouldn’t Be This Good A Caltech startup just crammed 8.2 billion parameters into 1.15 GB. I benchmarked it against Llama 3.1, Qwen 3, and Gemma 4 — and the…

levelup.gitconnected.com/i-tested-the...

#PrismML 8.2 billion parameters in 1.15 GB competes with #Llama3.1, #Qwen3, and #Gemma4 FP16 models in16 GB. PrismML’s Bonsai 8B is 14x smaller. On iPhone 17 Pro Max, it clocks 44 tokens per second: real-time conversation speed on a phone, no cloud required.

Alibaba Unveils Qwen3.6-Plus AI Model Alibaba unveiled Qwen3.6-Plus on Apr 2, 2026; launch targets agentic AI and enterprise pilots, with commercial rollouts expected in 2H 2026 and potential cloud revenue implications.

Alibaba Unveils Qwen3.6-Plus AI Model: Alibaba unveiled Qwen3.6-Plus on Apr 2, 2026; launch targets agentic AI and enterprise pilots, with commercial rollouts expected in 2H 2026 and potential cloud revenue implications. 👈 Read full analysis #Alibaba #Qwen3 #AI #AgenticAI #EnterpriseAI

Using #ollama with #Qwen3 (30b) to finally sort through the ~10K unread emails.

After it's done, I'll start reading them as they come, pinky swear! (I probably won't though, I know myself)

#ai #UsefulAI

Qwen3.5-35B-A3B Uncensored Guide: Features, Capabilities, and Setup

Explore Qwen3.5-35B-A3B Uncensored, a multimodal AI model with no refusals, advanced reasoning, and support for local deployment. #qwen3.5model

ansible/README.md at master ansible

Running a local #qwen3 model over #opencode produced the below vibe coding results - an Ansible reference. It's not much but it was honest work :) git.wtf.lt/simonas/ansible/src/bran...

Petal für macOS: Simple Diktat-App für die Menüleiste mit Qwen, Parakeet und Voxtral Petal ist eine native macOS-Menüleisten-App für schnelle, lokale Audio-Transkription. Die quelloffene App nutzt wahlweise die native Diktierfunktion von macOS oder Modelle von Qwen, FluidAudio, Whisper und Voxtral. Seit KI-Sprachmodelle ihren Siegeszug angetreten haben, ist auch das Thema der Audio-Transkription wieder in den Fokus gerückt. In der Vergangenheit waren Speech-to-Text-Anwendungen eher mühselig und deren Texte ohne…

Petal für #macOS: Simple Diktat-App für die Menüleiste mit #Qwen3, #Parakeet und #Voxtral. #OpenSource, ohne Cloud und Abo 🎙️🖥️🔓 https://tchgdns.de/?p=162724

Qwen 3 vs OpenAI & DeepSeek: In-Depth Technical Comparison for API Developers Qwen 3, Alibaba's next-gen language model, challenges OpenAI and DeepSeek with its scalable Mixture-of-Experts architecture, open weights, and robust benchmark results. This technical comparison dives deep into Qwen 3’s strengths for API and backend developers.

【話題】Qwen 3 が OpenAI や DeepSeek を凌駕する場面も？

Qwen 3 の進化速度が凄まじいですね。推論能力だけでなく、特定のタスクでの効率性でトップモデルに食い込む結果が出ています。皆さんは最近、どのモデルをメインにコード書いてますか？💡

#AI #LLM #Qwen3 #プログラミング

apidog.com/blog/qwen-3-outcompetes-...

【ローカルLLM】Qwen3を使ってみた話【中国最新AI】｜Catapp-Art3D はじめに Ollamaで大規模言語モデルQwen3が利用可能になったので利用してみました。日本語も公式対応しています。 ただし、筆者の主力GPUであるRTX 3060はFramepack「叡智」版のテストに利用しているので、GTX 1660ti+1650上でのQwen3 8B版の動作確認です。 Qwen3: Think Deeper, Act Faster QWEN CHAT GitHub Hugging Face ModelScope Kaggle DEMO DISCORD qwenlm.github.io 上記ページをQwen3 8B版で日本語要約し

🇨🇳【中国 AI「Qwen3」が凄すぎる】ローカルで動かせる 8B モデルが日本語・論理推論で神性能😱 Ollama で「ollama pull qwen3:8b」だけ！教室の席順パズルを 1 発で正解、9 千文字の日本史要約も完璧✨ 119 言語対応・Apache2.0 で商用自由。日本語チューニングモデル超え？ローカル AI 新時代ですね。Qwen3 試したことありますか？ #AI #ローカルLLM #Qwen3 https://note.com/catap_art3d/n/n1b486fc699f8

Snowflake's Arctic Long Sequence Training: How to Train LLMs on 15 Million Tokens Without Selling a Kidney Snowflake AI Research just open-sourced Arctic Long Sequence Training (ALST), a framework that pushes LLM training from a measly 32K tokens to over 15 million — a 469x improvement — using standard Hug...

Snowflake's Arctic Long Sequence Training: How to Train LLMs on 15 Million Tokens Without Selling a Kidney

techlife.blog/posts/snowfl...

#ALST #Snowflake #LongContextTraining #DeepSpeed #HuggingFace #SequenceParallelism #LLMTraining #H100 #Llama8B #Qwen3 #GPUMemoryOptimization

Screenshot of a dark-themed billing dashboard titled “March 2026” showing GPU usage charges totaling **$102.67**. Two entries are listed for **gpu_1x_a100_sxm4** instances launched by a **redacted email address** in **Arizona, USA**. The first job (“steeve trainer”) ran from **2026-03-05 01:27 UTC to 2026-03-07 21:58 UTC** for **68.52 hours** at **$1.48/hr**, costing **$101.41**. The second ran from **2026-03-05 00:29 UTC to 2026-03-05 01:21 UTC** for **0.86 hours**, costing **$1.26**. Each row includes a “Copy ID” link.

Advertisement-style image set in an office lobby showing a large display with a portrait of a middle-aged man in a suit. Beside the screen, bold text reads “STEEVE VERSION 6” with a smaller green digital-style sign below stating “NOW AVAILABLE.”

#Steeve 6 just dropped. $100 to train. 100k lines of #LoRA training on top of #Qwen 3.5. Vision now in-built so he really *can* see. Not to brag or anything, but let's see #OpenAI pull this off! :steeve:

#llm #chatbot #qwen3

Собственная облачная LLM на 16 ГБ VRAM — часть 1: базовая сборка, tools и MCP Привет, Хабр! На фоне ажиотажа вокруг ней...

#langchain #langgraph #python #vllm #qwen3 #localai #selectel #MCP #ии-агенты #API-сервис

Origin | Interest | Match

Heaven help me, I'm trying to re-train #Steeve from a new foundational model. Again. :steeve:

#qwen3 #qwen #ai #llm #chatbot

Alibaba just shocked the AI world.

And almost nobody is talking about it.

They dropped 4 new models called Qwen 3.5 Medium Model Series.

106,000 downloads in 12 hours.

#LearnAI #AutomateWithJulian #OpenSourceAI #Qwen3

Der Schweizer Chatbot Euria verbreitet die Ideologie der chinesischen Regierung Die Schweizer Cloud-Firma Infomaniak preist ihren Chatbot als lokale Alternative zu ausländischen KI-Systemen wie Chat-GPT an. Doch Euria basiert auf einem chinesischen Modell – und verhält sich entsp...

#Infomaniak wirbt aggressiv mit digitaler Unabhängigkeit, Ethik und «Swissness», verwendet aber chinesische Qwen3-KI-Modell … 😬

www.nzz.ch/nzz-am-sonnt...

#ai #ki #qwen3 #china #schweiz

The open source AI world just caught up to ChatGPT and Claude — and you can run the whole thing for free.

Pair it with Ollama and OpenClaw and it becomes a full agent. Token limits reset weekly and most people aren't even hitting them.

#OpenSourceAI #FreeAI #AITools #Qwen3 #OpenClaw

Чат-бот с генеративным AI под капотом (разговаривает как живой) Реализация чат-бота с LLM под капотом, который ...

#чат-бот #qwen #qwen3 #qwen2.5

Origin | Interest | Match

#AItutorial #Qwen3 #ASR #stt #karaoke #diy

youtu.be/Dhn_7VTDBhc

Here the comparison between Cloud API and local Qwen. But still impressive how reliable #Qwen3 is.

After testing various small local LLMs for tool-calling capabilities, the winner is: #Qwen3 8B. It's very reliable but also kinda slow in a sovereign setup. The native tool calling works beautifully though. Love it.

[Shoper Gamer] Alibaba Cloud เปิดตัวโมเดล AI เขียนโค้ด “Qwen3-Coder-Next” ขนาดเล็ก-รันเร็ว-ราคาประหยัด โดย โดย

Alibaba Cloud เปิดตัวโมเดล AI เขียนโค้ด “Qwen3-Coder-Next” ขนาดเล็ก-รันเร็ว-ราคาประหยัด

#ShoperGamer #Qwen #Alibaba #AlibabaCloud #Ai #ModelAi #Qwen3 #Qwen3CoderNext #Feed

Alibaba’s Qwen3-Coder-Next Activates Just 3B of 80B Parameters For Improved Efficiency Alibaba has released Qwen3-Coder-Next, an open-source 80B-parameter coding model that activates just 3B parameters per query, scoring 70.6% on SWE-Bench.

winbuzzer.com/2026/02/04/a...

Alibaba’s Qwen3-Coder-Next Activates Just 3B of 80B Parameters For Improved Efficiency

#AI #AICoding #Qwen3 #Qwen3CoderNext #Alibaba #MixtureOfExperts LargeLanguageModels #OpenSourceAI #Coding

¡Llegó Qwen3-Max-Thinking! El nuevo gigante de la #IA que compite con GPT-5.2 y Claude-Opus-4.5. Destaca por su razonamiento avanzado, uso autónomo de herramientas y máxima eficiencia en tareas complejas. 🤖✨
youtu.be/xoWwCzeJrY8
#Qwen3 #IA #TechNews #SEO #MachineLearning

Alibaba Cloud’s Qwen-3 becomes the world’s first AI model to operate in orbit

Build your freelance career smarter, faster, better — Join Realancer 👉 realancer.net

🔗 Read full blog here: alitech.io/blog/qwen-3-...

#AI #ArtificialIntelligence #Qwen3 #AlibabaCloud #SpaceAI #Realancer #Viral

Amazon (company) - Wikipedia

Here are the sources linked at the bottom, since I couldn't get them to fit in the generated alt text (using #qwen3 locally, in #LMStudio, btw), even after paring it down:
www.amazon.com (although you probably won't click that if you're trying to boycott them), en.wikipedia.org/wiki/Amazon_...

#Qwen3 #MaxThinking is a new #reasoning model that achieves performance comparable to leading models like GPT-5.2-Thinking and Claude-Opus-4.5. It features adaptive tool-use capabilities, allowing it to autonomously select and leverage Search, Memory, and Code Interpreter tools during…

[Audio] Original post on maly.io

🦜

https://huggingface.co/spaces/Qwen/Qwen3-TTS

#TTS #Qwen3

Dev Weekly #95 – Qwen3-TTS Family is Now Open Sourced, jQuery Turns 20 & Modernizes, Astro Powers Up with Cloudflare - The Miners Hello everyone, the latest Dev Weekly #95 is out! Qwen3-TTS Family is Now Open Sourced, jQuery 4.0, Astro joins Cloudflare & more hot tech updates. Read now!

Friday's here, and you know what that means -- time for Dev Weekly!

This week's tech landscape has been absolutely buzzing, from breakthrough releases to industry shake-ups, check it out!

bit.ly/3LQ1Iow

#Codeminer42 #Qwen3 #jQuery #Astro #Cloudflare #Flutter #AI #RubyOnRails

Running and deploying LLMs like Qwen3-0.6B on your Android or iPhone. A truly smart smartphone with the same technology Meta uses to power billions of users on Instagram and WhatsApp. #UnslothAI #Qwen3 #iOS #Android #AI #Qwen #OpenSource #Tech

ℹ️ tinyurl.com/yyeak578

¡Alibaba rompe récords con Qwen3-VL!
El nuevo estándar SOTA en búsqueda multimodal que unifica texto, 🖼️ imágenes y 🎥 video en un solo sistema.
✅ #1 en MMEB-V2 (score 77.8) ✅ Soporta +30 idiomas ✅ Eficiencia extrema con MRL
youtu.be/zgrSDSTPfzI
#Qwen3 #MultimodalAI #SEO