#localai — Public Fediverse posts

Hi Local AI enthusiasts! Just wanted to share a neat discovery I made while experimenting with local AI setups.

I’ve been a long time user of the Obsidian note taking app and recently discovered this Co-pilot plugin, it’s pretty cool! It lets you connect your own local models directly to your notes.

Before I would ask LM Studio → copy/paste → Obsidian and now I can ask questions about ANY note right IN Obsidian, and I can create content and get summaries inserted instantly! 😊

Anyone else experimenting with local LLMs and/or Obsidian lately? What else am I missing out on?! Would love to hear what you're running!

Check out Obsidian Co-Pilot: https://github.com/logancyang/obsidian-copilot

#localai #opensource #privacyfirst #obsidian #fediverse #tinker #mastodon #ai #buildinpublic #tech

Hi Local AI enthusiasts! Just wanted to share a neat discovery I made while experimenting with local AI setups.

I’ve been a long time user of the Obsidian note taking app and recently discovered this Co-pilot plugin, it’s pretty cool! It lets you connect your own local models directly to your notes.

Before I would ask LM Studio → copy/paste → Obsidian and now I can ask questions about ANY note right IN Obsidian, and I can create content and get summaries inserted instantly! 😊

Anyone else experimenting with local LLMs and/or Obsidian lately? What else am I missing out on?! Would love to hear what you're running!

Check out Obsidian Co-Pilot: https://github.com/logancyang/obsidian-copilot

#localai #opensource #privacyfirst #obsidian #fediverse #tinker #mastodon #ai #buildinpublic #tech

New week, more slides: Run LLMs Locally

Now including wllama to run GGUF models inside your browser!

wllama uses llama.cpp, WebAssembly and WebGPU, bringing a completely new experience of LLMs into the web.
It has no 4 GB limitation and is faster than Transformers.js.

I also added translations using the HY-MT model from Tencent.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #wllama #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp #webassembly

New week, more slides: Run LLMs Locally

Now including wllama to run GGUF models inside your browser!

wllama uses llama.cpp, WebAssembly and WebGPU, bringing a completely new experience of LLMs into the web.
It has no 4 GB limitation and is faster than Transformers.js.

I also added translations using the HY-MT model from Tencent.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #wllama #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp #webassembly

New week, more slides: Run LLMs Locally

Now including wllama to run GGUF models inside your browser!

wllama uses llama.cpp, WebAssembly and WebGPU, bringing a completely new experience of LLMs into the web.
It has no 4 GB limitation and is faster than Transformers.js.

I also added translations using the HY-MT model from Tencent.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #wllama #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp #webassembly

New week, more slides: Run LLMs Locally

Now including wllama to run GGUF models inside your browser!

wllama uses llama.cpp, WebAssembly and WebGPU, bringing a completely new experience of LLMs into the web.
It has no 4 GB limitation and is faster than Transformers.js.

I also added translations using the HY-MT model from Tencent.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #wllama #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp #webassembly

New week, more slides: Run LLMs Locally

Now including wllama to run GGUF models inside your browser!

wllama uses llama.cpp, WebAssembly and WebGPU, bringing a completely new experience of LLMs into the web.
It has no 4 GB limitation and is faster than Transformers.js.

I also added translations using the HY-MT model from Tencent.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #wllama #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp #webassembly

I got a successful tool call running in Pi Coding Agent with local Ollama qwen3.5:9b.

Building local-agent-bench on 8GB VRAM to test how local LLMs handle agent tasks: tool calls, tool results, multi-step workflows, and hallucination checks.

The interesting part: Pi works here, while OpenClaw / Hermes native tool calls do not yet. That gives me a clean benchmark target.

#LocalAI #Ollama #LLM #openclaw #hermes #pi #qwen #ai

RT @googlegemma: Wir betreten eine neue Ära der On-Device-Automatisierung. ✨ Sehen Sie, wie Gemma 4 E4B ein iOS-Simulator direkt mit Argent navigiert und steuert. Lokale Modelle können komplexe Interaktionen und Software-Navigation autonom bewältigen. Video

mehr auf Arint.info

#AutonomousDriving #Gemma4 #iOSAutomation #LocalAI #OnDeviceAutomation #TechInnovation #arint_info

https://x.com/googlegemma/status/2057570113390551452#m

RT @TeksEdge: 🔥 Offizielle Updates zu AMDs Ryzen AI Halo PC

mehr auf Arint.info

#AMD #GamingPC #KIHardware #LocalAI #RyzenAI #StrixHalo #arint_info

https://x.com/TeksEdge/status/2058063559767241211#m

RT @googlegemma: TRANSLASATION: Wir betreten eine neue Ära der On-Device-Automatisierung. ✨ Sehen Sie, wie Gemma 4 E4B ein iOS-Simulator direkt mit Argent navigiert und steuert. Lokale Modelle können komplexe Interaktionen und Software-Navigation autonom bewältigen. Video

mehr auf Arint.info

#Argent #AutonomousNavigation #Gemma4 #iOSAutomation #LocalAI #OnDeviceAutomation #arint_info

https://x.com/googlegemma/status/2057570113390551452#m

Lokale #KI suckt auf #RasperryPi. Wer #squatted mit mir ein paar #DataCenter? #Ollama #Localai #DockerModelRunner #LMStudio #ComfyUI

Lokale #KI suckt auf #RasperryPi. Wer #squatted mit mir ein paar #DataCenter? #Ollama #Localai #DockerModelRunner #LMStudio #ComfyUI

Lokale #KI suckt auf #RasperryPi. Wer #squatted mit mir ein paar #DataCenter? #Ollama #Localai #DockerModelRunner #LMStudio #ComfyUI

Unlock WhatsApp Data with Local Analytics Dashboard

Most people think of WhatsApp as “just messaging.”

But after years of conversations, support threads, customer discussions, team coordination, and random life moments… it quietly becomes one of the richest personal datasets you own.

So I built wacrawl-ui — a local analytics dashboard for WhatsApp archives generated by wacrawl.

The idea is simple:

• Your data stays local
• No cloud sync
• No browser extension
• No scraping APIs
• No “AI magic” uploading your chats somewhere

Just a fast local dashboard on top of SQLite.

What’s inside:

• Full-text search (FTS5) – It’s working quite fast. Even on ~100k messages.
• Messaging activity analytics
• Contact insights
• Media browsing
• Response-time patterns
• Word clouds
• Group activity stats
• Read-only local API
• React + Vite frontend
• Express backend
• Zero external dependencies once running – You only need to make sure you run ‘wacrawl sync‘ before.

A few things I found interesting while building it:

1. SQLite is still absurdly powerful
   People underestimate what you can do locally with FTS indexes and good schema design.
2. “Local-first” UX matters more than ever
   We’ve normalized uploading deeply personal data to random SaaS products. We should challenge that assumption.
3. Personal analytics is an untapped category
   Not surveillance. Not ad targeting.
   Tools that help you understand your own data.
4. Read-only architectures reduce risk dramatically
   The app intentionally avoids mutation flows. That constraint simplified security and reliability decisions across the stack.

The whole thing runs with one line:

 npx wacrawl-dashboard@latest 

No complicated setup.

Still early, but I think there’s a broader shift happening toward:

• local AI – Ollama for the win.
• local analytics – secure, private and handy.
• local search – that works fast.
• user-owned datasets – It’s not for everyone, but it’s useful.

Well, that future feels healthier.

Feel free to check the repo: github.com/greenido/wacrawl-ui and contribute.

Be strong.

Rate this:

Unlock WhatsApp Data with Local Analytics Dashboard

Most people think of WhatsApp as “just messaging.”

But after years of conversations, support threads, customer discussions, team coordination, and random life moments… it quietly becomes one of the richest personal datasets you own.

So I built wacrawl-ui — a local analytics dashboard for WhatsApp archives generated by wacrawl.

The idea is simple:

• Your data stays local
• No cloud sync
• No browser extension
• No scraping APIs
• No “AI magic” uploading your chats somewhere

Just a fast local dashboard on top of SQLite.

What’s inside:

• Full-text search (FTS5) – It’s working quite fast. Even on ~100k messages.
• Messaging activity analytics
• Contact insights
• Media browsing
• Response-time patterns
• Word clouds
• Group activity stats
• Read-only local API
• React + Vite frontend
• Express backend
• Zero external dependencies once running – You only need to make sure you run ‘wacrawl sync‘ before.

A few things I found interesting while building it:

1. SQLite is still absurdly powerful
   People underestimate what you can do locally with FTS indexes and good schema design.
2. “Local-first” UX matters more than ever
   We’ve normalized uploading deeply personal data to random SaaS products. We should challenge that assumption.
3. Personal analytics is an untapped category
   Not surveillance. Not ad targeting.
   Tools that help you understand your own data.
4. Read-only architectures reduce risk dramatically
   The app intentionally avoids mutation flows. That constraint simplified security and reliability decisions across the stack.

The whole thing runs with one line:

 npx wacrawl-dashboard@latest 

No complicated setup.

Still early, but I think there’s a broader shift happening toward:

• local AI – Ollama for the win.
• local analytics – secure, private and handy.
• local search – that works fast.
• user-owned datasets – It’s not for everyone, but it’s useful.

Well, that future feels healthier.

Feel free to check the repo: github.com/greenido/wacrawl-ui and contribute.

Be strong.

Unlock WhatsApp Data with Local Analytics Dashboard

Most people think of WhatsApp as “just messaging.”

But after years of conversations, support threads, customer discussions, team coordination, and random life moments… it quietly becomes one of the richest personal datasets you own.

So I built wacrawl-ui — a local analytics dashboard for WhatsApp archives generated by wacrawl.

The idea is simple:

• Your data stays local
• No cloud sync
• No browser extension
• No scraping APIs
• No “AI magic” uploading your chats somewhere

Just a fast local dashboard on top of SQLite.

What’s inside:

• Full-text search (FTS5) – It’s working quite fast. Even on ~100k messages.
• Messaging activity analytics
• Contact insights
• Media browsing
• Response-time patterns
• Word clouds
• Group activity stats
• Read-only local API
• React + Vite frontend
• Express backend
• Zero external dependencies once running – You only need to make sure you run ‘wacrawl sync‘ before.

A few things I found interesting while building it:

1. SQLite is still absurdly powerful
   People underestimate what you can do locally with FTS indexes and good schema design.
2. “Local-first” UX matters more than ever
   We’ve normalized uploading deeply personal data to random SaaS products. We should challenge that assumption.
3. Personal analytics is an untapped category
   Not surveillance. Not ad targeting.
   Tools that help you understand your own data.
4. Read-only architectures reduce risk dramatically
   The app intentionally avoids mutation flows. That constraint simplified security and reliability decisions across the stack.

The whole thing runs with one line:

 npx wacrawl-dashboard@latest 

No complicated setup.

Still early, but I think there’s a broader shift happening toward:

• local AI – Ollama for the win.
• local analytics – secure, private and handy.
• local search – that works fast.
• user-owned datasets – It’s not for everyone, but it’s useful.

Well, that future feels healthier.

Feel free to check the repo: github.com/greenido/wacrawl-ui and contribute.

Be strong.

Rate this:

Unlock WhatsApp Data with Local Analytics Dashboard

Most people think of WhatsApp as “just messaging.”

But after years of conversations, support threads, customer discussions, team coordination, and random life moments… it quietly becomes one of the richest personal datasets you own.

So I built wacrawl-ui — a local analytics dashboard for WhatsApp archives generated by wacrawl.

The idea is simple:

• Your data stays local
• No cloud sync
• No browser extension
• No scraping APIs
• No “AI magic” uploading your chats somewhere

Just a fast local dashboard on top of SQLite.

What’s inside:

• Full-text search (FTS5) – It’s working quite fast. Even on ~100k messages.
• Messaging activity analytics
• Contact insights
• Media browsing
• Response-time patterns
• Word clouds
• Group activity stats
• Read-only local API
• React + Vite frontend
• Express backend
• Zero external dependencies once running – You only need to make sure you run ‘wacrawl sync‘ before.

A few things I found interesting while building it:

1. SQLite is still absurdly powerful
   People underestimate what you can do locally with FTS indexes and good schema design.
2. “Local-first” UX matters more than ever
   We’ve normalized uploading deeply personal data to random SaaS products. We should challenge that assumption.
3. Personal analytics is an untapped category
   Not surveillance. Not ad targeting.
   Tools that help you understand your own data.
4. Read-only architectures reduce risk dramatically
   The app intentionally avoids mutation flows. That constraint simplified security and reliability decisions across the stack.

The whole thing runs with one line:

 npx wacrawl-dashboard@latest 

No complicated setup.

Still early, but I think there’s a broader shift happening toward:

• local AI – Ollama for the win.
• local analytics – secure, private and handy.
• local search – that works fast.
• user-owned datasets – It’s not for everyone, but it’s useful.

Well, that future feels healthier.

Feel free to check the repo: github.com/greenido/wacrawl-ui and contribute.

Be strong.

Rate this:

OWC Stack AI brings Thunderbolt 5 local AI support to Windows and Linux

https://fed.brid.gy/r/https://nerds.xyz/2026/05/owc-stack-ai-thunderbolt-5/

OWC Stack AI brings Thunderbolt 5 local AI support to Windows and Linux

https://web.brid.gy/r/https://nerds.xyz/2026/05/owc-stack-ai-thunderbolt-5/

OWC Stack AI brings Thunderbolt 5 local AI support to Windows and Linux

https://web.brid.gy/r/https://nerds.xyz/2026/05/owc-stack-ai-thunderbolt-5/

OWC Stack AI brings Thunderbolt 5 local AI support to Windows and Linux

https://fed.brid.gy/r/https://nerds.xyz/2026/05/owc-stack-ai-thunderbolt-5/

OWC Stack AI brings Thunderbolt 5 local AI support to Windows and Linux

https://web.brid.gy/r/https://nerds.xyz/2026/05/owc-stack-ai-thunderbolt-5/

Running Local AI on Windows: A Technocratic Hurdle

Running AI locally on Windows with a GPU needs special tech setup. Users need to install WSL and compatible drivers. Learn why it matters.

#LocalAI, #WindowsTech, #GPUSetup, #WSL, #AICoding

https://newsletter.tf/local-ai-windows-gpu-tech-setup/

Setting up AI on your own Windows computer is harder than you think. It requires installing special software like WSL and making sure your graphics card works with it.

#LocalAI, #WindowsTech, #GPUSetup, #WSL, #AICoding
https://newsletter.tf/local-ai-windows-gpu-tech-setup/

I wonder if using a "dumber" local AI model might help mitigate the cognitive decline some researchers are starting to observe.

I am currently running Qwen 3.6 26B via Llama.cpp and it's obviously not as good as Claude or Gemini. It requires some hand-holding. But that doesn't mean it's useless. It can still provide insights, but it's up to you to steer it to get you the results you desire.

#llm #ai #qwen #localai

I wonder if using a "dumber" local AI model might help mitigate the cognitive decline some researchers are starting to observe.

I am currently running Qwen 3.6 26B via Llama.cpp and it's obviously not as good as Claude or Gemini. It requires some hand-holding. But that doesn't mean it's useless. It can still provide insights, but it's up to you to steer it to get you the results you desire.

#llm #ai #qwen #localai

I wonder if using a "dumber" local AI model might help mitigate the cognitive decline some researchers are starting to observe.

I am currently running Qwen 3.6 26B via Llama.cpp and it's obviously not as good as Claude or Gemini. It requires some hand-holding. But that doesn't mean it's useless. It can still provide insights, but it's up to you to steer it to get you the results you desire.

#llm #ai #qwen #localai

I wonder if using a "dumber" local AI model might help mitigate the cognitive decline some researchers are starting to observe.

I am currently running Qwen 3.6 26B via Llama.cpp and it's obviously not as good as Claude or Gemini. It requires some hand-holding. But that doesn't mean it's useless. It can still provide insights, but it's up to you to steer it to get you the results you desire.

#llm #ai #qwen #localai

I wonder if using a "dumber" local AI model might help mitigate the cognitive decline some researchers are starting to observe.

I am currently running Qwen 3.6 26B via Llama.cpp and it's obviously not as good as Claude or Gemini. It requires some hand-holding. But that doesn't mean it's useless. It can still provide insights, but it's up to you to steer it to get you the results you desire.

#llm #ai #qwen #localai

New week, new slides: Run LLMs Locally

Now including multi-token prediction using Qwen3.6 35B-A3B with Nextn quantization. Also speech recognition using Qwen-3-ASR is now working directly with Llama.cpp and included in the slides.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp

New week, new slides: Run LLMs Locally

Now including multi-token prediction using Qwen3.6 35B-A3B with Nextn quantization. Also speech recognition using Qwen-3-ASR is now working directly with Llama.cpp and included in the slides.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp

New week, new slides: Run LLMs Locally

Now including multi-token prediction using Qwen3.6 35B-A3B with Nextn quantization. Also speech recognition using Qwen-3-ASR is now working directly with Llama.cpp and included in the slides.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp

New week, new slides: Run LLMs Locally

Now including multi-token prediction using Qwen3.6 35B-A3B with Nextn quantization. Also speech recognition using Qwen-3-ASR is now working directly with Llama.cpp and included in the slides.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp

New week, new slides: Run LLMs Locally

Now including multi-token prediction using Qwen3.6 35B-A3B with Nextn quantization. Also speech recognition using Qwen-3-ASR is now working directly with Llama.cpp and included in the slides.

https://codeberg.org/thbley/talks/raw/branch/main/Run_LLMs_Locally_2026_ThomasBley.pdf

#ai #llm #llamacpp #stablediffusion #qwen3 #glm #localai #gemma4 #webgpu #opencode #mtp

I just built a Discord bot powered by a local AI model. No API keys. No cloud. Runs entirely on your machine. https://medium.com/p/e3e43703d95d
#Discord #Python #Ollama #AI #LocalAI

Reddit Developer Unlocks Multi-GPU Power With NVENC

How does the new ATLAS software allow $500 multi-GPU setups to match cloud LLM performance? Learn how this tool helps developers run coding AI locally.

#localai, #gpu, #coding, #nvidia, #techupdate

https://newsletter.tf/reddit-atlas-multi-gpu-llm-coding-boost/

A new software project called ATLAS lets users link multiple GPUs to run coding AI. This $500 setup now performs as well as expensive cloud services like Claude Sonnet.

#localai, #gpu, #coding, #nvidia, #techupdate
https://newsletter.tf/reddit-atlas-multi-gpu-llm-coding-boost/

Aylık Aboneliklere Elveda: Kendi Bilgisayarınızda Bedava Yerel AI

Claude ve ChatGPT'ye her ay dolarla abonelik ücreti ödemek istemeyenler için kendi donanımımız üzerinde çalışan gizli ve bedava AI ekosistemi kuruyoruz. GPU VRAM yönetimi ilkelerini, LM Studio ile model sıkıştırmayı (Quantization - Q4/Q6), VS Code Continue eklentisiyle lokal Copilot entegrasyonunu ve Qwen 2.5 ile yaptığım Sudoku benchmark testlerini detaylıca anlattım.

https://yuceltoluyag.github.io/yerel-ai-kurulum-rehberi-lm-studio-vs-code/

#localai #lmstudio #vscode #llm #ai

Local AI Emergence: Hermes Agent Navigates New Frontiers

Hermes Agent lets you run AI models like Gemma 4 and Qwen 3.5 on your own computer. Find out how it works and why it matters for AI users.

#LocalAI, #HermesAgent, #LLM, #OpenSourceAI, #TechNews

https://newsletter.tf/run-ai-locally-hermes-agent-gemma-qwen/

You can now run advanced AI models on your computer with Hermes Agent. This is a big change from needing expensive cloud services.

#LocalAI, #HermesAgent, #LLM, #OpenSourceAI, #TechNews
https://newsletter.tf/run-ai-locally-hermes-agent-gemma-qwen/

RT @stevibe: Parameter-Scaling ist gerade bei mir zusammengebrochen. Ich habe 90 Matheaufgaben als Bilder an 10 lokale Vision-Modelle gegeben, jeweils 3 Durchläufe pro Modell, wobei nur konsistente Antworten über alle 3 Durchläufe gezählt wurden. Zwei Erkenntnisse: Gemma 4 war die konsistenteste Familie, 31B holte sich den Titel mit 89,6%. Doch Qwen 3.5 4B lag nur 2 Antworten dahinter. Ein 4B-Modell. Auf Platz 2 von 10. Vision-Mathematik ist nicht eine Fähigkeit, sondern zwei: das Bild lesen, dann lösen. Die eigentliche Lektion für alle, die lokal arbeiten: klein ≠ schwach. Wenn du agentic Workflows aufbaust, ist es wichtiger, das richtige Modell für jede Aufgabe zu finden, als zum größten Modell zu greifen. In diesem Test lief das 4B-Modell aufgrund seiner Größe weit schneller, erzielte höhere Scores und ließ VRAM für den Rest deines Stacks frei. Vollständige Ergebnisse: 🥇 Gemma 4 31B — 242/270 (89,6%) 🥈 Qwen 3.5 4B — 240/270 (88,9%) 🥉 Gemma 4 E4B — 222/270 (82,2%) 🥉 Qwen 3.6 27B — 222/270 (82,2%) 5. Gemma 4 26B A4B — 216/270 (80,0%) 6. Qwen 3.5 2B — 201/270 (74,4%) 7. Gemma 4 E2B — 192/270 (71,1%) 8. Qwen 3.6 35B A3B — 192/270 (71,1%) 9. Qwen 3.5 9B — 168/270 (62,2%) 10. Qwen 3.5 0.8B — 45/270 (16,7%) Alle GGUF + mmproj, Unsloth's Q6KXL Quantisierung. Video

mehr auf Arint.info

#Gemma #LLM #LocalAI #MachineLearning #Qwen #VisionModels #arint_info

https://x.com/stevibe/status/2055666729460932626#m

- Junyang Lin, ex Qwen lead starts new lab: https://www.theinformation.com/articles/former-alibaba-star-researcher-starts-new-ai-lab-seeks-2-billion-valuation

- Google DeepMind AI (mouse) pointer research makes sense for XR: https://deepmind.google/blog/ai-pointer/

- HRF awards grant to 0xSero to run open models locally for users who face regime surveillance: https://hrf.org/latest/hrfs-ai-fund-supports-10-innovative-projects/

#AI #AINews #Qwen #googledeepmind #xr #opensource #localai

- Junyang Lin, ex Qwen lead starts new lab: https://www.theinformation.com/articles/former-alibaba-star-researcher-starts-new-ai-lab-seeks-2-billion-valuation

- Google DeepMind AI (mouse) pointer research makes sense for XR: https://deepmind.google/blog/ai-pointer/

- HRF awards grant to 0xSero to run open models locally for users who face regime surveillance: https://hrf.org/latest/hrfs-ai-fund-supports-10-innovative-projects/

#AI #AINews #Qwen #googledeepmind #xr #opensource #localai

- Junyang Lin, ex Qwen lead starts new lab: https://www.theinformation.com/articles/former-alibaba-star-researcher-starts-new-ai-lab-seeks-2-billion-valuation

- Google DeepMind AI (mouse) pointer research makes sense for XR: https://deepmind.google/blog/ai-pointer/

- HRF awards grant to 0xSero to run open models locally for users who face regime surveillance: https://hrf.org/latest/hrfs-ai-fund-supports-10-innovative-projects/

#AI #AINews #Qwen #googledeepmind #xr #opensource #localai

- Junyang Lin, ex Qwen lead starts new lab: https://www.theinformation.com/articles/former-alibaba-star-researcher-starts-new-ai-lab-seeks-2-billion-valuation

- Google DeepMind AI (mouse) pointer research makes sense for XR: https://deepmind.google/blog/ai-pointer/

- HRF awards grant to 0xSero to run open models locally for users who face regime surveillance: https://hrf.org/latest/hrfs-ai-fund-supports-10-innovative-projects/

#AI #AINews #Qwen #googledeepmind #xr #opensource #localai

- Junyang Lin, ex Qwen lead starts new lab: https://www.theinformation.com/articles/former-alibaba-star-researcher-starts-new-ai-lab-seeks-2-billion-valuation

- Google DeepMind AI (mouse) pointer research makes sense for XR: https://deepmind.google/blog/ai-pointer/

- HRF awards grant to 0xSero to run open models locally for users who face regime surveillance: https://hrf.org/latest/hrfs-ai-fund-supports-10-innovative-projects/

#AI #AINews #Qwen #googledeepmind #xr #opensource #localai

...Dinoki’s customers had asked him why they should buy the app if they still had to pay for tokens — the usage units AI companies charge for processing prompts and generating responses...

Local/centralized hybrid. Apple silicon only. For now.

https://techcrunch.com/2026/05/15/osaurus-brings-both-local-and-cloud-ai-models-to-your-mac/

#localai #ai #Apple

🤖 janhq/jan

Jan is an open source alternative to ChatGPT that runs 100% offline on your computer.

Runs open-source LLMs like Llama or GPT alternatives entirely offline with local model downloads, cloud API integrations and a privacy-focused desktop app for Windows, macOS and Linux

⭐ Stars: 42541
📅 Last Update: May 15, 2026

https://github.com/janhq/jan

#selfhosted #homelab #selfhost #selfhosting #opensource #localai #offlinellm