In den vergangenen Monaten häufen sich auch bei Tails die Notfall-Releases aufgrund von durch KI-Tools entdeckter Sicherheitslücken. Tails 7.8.1 schließt einen weiteren Linux Kernel Privilege Escalation Bug.

Ever wished you could talk in to a text field rather than type? Ubuntu 26.10 hears you – quite literally. Canonical’s VP of Engineer Jon Seager, at the Ubuntu Summit, said the distro will soon lets users “press a button and talk into any field that you could previously type in”. A small, on-device AI language parsing model like Whisper will power the feature. It’s part of a wider push to integrate AI features in Ubuntu this year, with founder Mark Shuttleworth aiming to position Ubuntu as the ‘OS for agentic AI’. AI features in Ubuntu will be shipped as […]

You're reading Ubuntu is adding AI-powered voice input to all text fields, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

IBM will mit Project Lightwell der rasant steigenden Zahl der durch KI aufgedeckten Sicherheitslücken Herr werden. Dazu sollen 5 Milliarden US-Dollar investiert werden.

I may hate AI slop but I am not a AI hater. I have found decent use of the AI tools and I try to include these tools in my workflow wherever it makes sense.

While mainstream LLMs like ChatGPT and Perplexity have decent free offering, they leach on the user data. "If you are not paying for the product, you are the product".

That's why I am loving the idea of exploring local AI and I have spend my fair share of time experimenting with LLMs that can be run on normal systems.

Recently, I discovered Jan AI. It is a polished, genuinely usable desktop app that runs entirely on my machine. In fact, I once tried replacing Ollama and llama cpp with Jan AI, but later changed my mind.

I'll explain why I switched back to Ollama in the later sections. First, let's learn about Jan AI.

What is Jan AI?

Jan is a free and open-source desktop application that lets you run various large language models directly on your own hardware. You can think of it as a self-hosted, offline-capable ChatGPT, except the model runs on your CPU or GPU, and no data ever leaves your machine.

The project is developed by the Jan.ai company, and the source code is available on GitHub under the AGPL-3.0 license. It's built on top of llama.cpp under the hood, which means it can run quantized GGUF models efficiently even without a dedicated GPU.

What I found impressive is that Jan's desktop application is built using the Tauri framework instead of Electron JS, which gives it a good performance boost and I think eats less RAM, too.

A quick catch, you still need plenty of RAM for running your local LLM.

The app supports Linux, macOS, and Windows. I used it on my Linux machine.

System requirements

Running a local LLM with Jan AI does require decent hardware. Here's what to realistically expect:

• 8 GB RAM minimum. It's enough for 7B parameter models at 4-bit quantization (Q4_K_M). You'll notice slowdowns with other applications open. 16 GB RAM is the sweet spot. It can comfortably run 7B models and lets you experiment with 13B models.
• GPU acceleration is optional but makes a big difference. Jan supports NVIDIA (via CUDA), AMD (via ROCm), and Intel Arc. If you don't have a compatible GPU, CPU-only mode works, but it's significantly slower.
• Disk space models range from around 4 GB (7B, Q4) to 8 GB+ (13B). Download only what you need.

Installing Jan AI on Linux

Jan offers multiple installation formats, including .deb and AppImage, but the AppImage is what I'd recommend for most Linux users. It's a single self-contained file that runs on virtually any distro without touching your system packages or requiring root privileges for the app itself. It requires no dependency hell. It's just made to work.

I am not going in the details but if you need help, refer to this guide to use AppImage on Linux.

When you run Jan AI for the first time, you should see an interface like this:

Downloading models and putting Jan to work

A freshly installed Jan is essentially an empty shell capable, but waiting. The first thing you need to do is grab a model. Think of it like the app is the frame and the model is the brain.

Finding your way to the model hub

Click the Hub in the left sidebar. This is Jan's built-in model library, a curated list of open-source models you can download with a single click.

There are dozens of models with names full of numbers and letters like Q4_K_M or IQ3_XS.

Before I recommend which ones to pick, let me quickly demystify that alphabet soup because I spent an embarrassing amount of time confused by it when I first started.

What does Q4_K_M actually mean?

Every model name carries a quantization tag. Quantization is how a model's original full-precision weights get compressed to fit on consumer hardware. Here's a practical breakdown:

• Q4: 4-bit quantization. The most common choice. Roughly 4 GB for a 7B model. Fast, memory-efficient, and quality holds up well for everyday tasks.
• Q8: 8-bit quantization. Nearly full quality, but needs about twice the RAM. Worth trying if you have 32 GB or more.
• K_M: The specific compression method (K-quants, medium variant). Generally, the best balance of speed and quality within the Q4 family.
• XS (Extra Small): This is the most aggressive compression. It results in a file size that is slightly smaller than a standard Q4_K_S

The 3 LLM models I tested

I am using an AMD laptop with an integrated GPU. I settled for CPU interference for local LLM models because enabling RoCm (Radeon Open Compute) is a hectic task, based on my research. It requires 25-30 GB of disk space.

For those of you who aren't familiar with what AMD ROCm is, it's the answer to NVIDIA's CUDA. Basically, when you run a local LLM, the heavy math, such as matrix multiplications across billions of parameters, can be offloaded to a GPU instead of the CPU.

RoCm lets software like Jan AI, PyTorch, or Llama.cpp talks to AMD Radeon GPUs to do that same heavy lifting. Without it, Jan falls back to CPU-only mode, which works but is significantly slower.

For this article, I settled on three models that cover a solid range of use cases. I asked the same question to each model - "What can you do"? Here's what I found:

Gemma 4

I don't lie. The performance wasn't superb, only 5 tokens per second.

Gemma 3

The performance was bad than the Gemma 4 2B model. It was only 3 tokens per second. I think you can use the Gemma 3 2B model variant to achieve comparable performance.

Jan

Yeah! Jan offers its own AI model. I used the Jan-v3-4b model to test. It comes with 262K context length. It can follow up instructions and do some light coding tasks.

Features I liked in Jan AI

Let's see a few features that I liked in this open source AI tool.

Unified Interface for Local and Cloud LLMs

Jan can act as a unified interface for local and cloud LLMs, instead of jumping from Claude to ChatGPT and then Gemini. You get to talk to your favorite LLMs within a single interface.

Though directly interacting with Cloud LLMs takes away your privacy, you can still use Jan AI as a unified interface to interact with any models on HuggingFace, Groq, Gemini, Qwen, Mistral, and Claude with the respective platform's API.

I'm especially in love with the Groq and Cerebras API. These platforms provide fast interference.

If your model provider is not in the list, you can still add OpenAI-compatible models to it.

To add, click on the plus icon next to "MODEL PROVIDERS". It'll ask for a name. Enter the name and then set the BASE_URL and API Key provided by your LLM platform.

Keeps your chats organized

Jan offers the functionality to organize chats per project.

Tailor Jan for Custom Experience

You can edit the system instruction. It's kind of making LLM behave like a specific expert.

Local API Server

It comes with a local API server. You can start the server by going to Settings -> Local API Server.

Use it as a tool to download AI models for Llama.cpp

Models downloaded with Jan AI application are completely compatible with Llama.cpp CLI. You can use Jan as a downloader tool and then run the downloaded AI model with llama-server a command.

Jan CLI

Jan offers a CLI interface to interact with as well.

My honest opinion on using Jan AI vs Ollama

Thanks to the developer involved in building the Jan AI app. It provides a nice GUI interface to interact with LLMs. On the contrary, I see some improvement scope for the application.

The major one is with the local LLM inference time. I tested the same model, and what I found was that Jan AI provided slow inference time as compared to Ollama on the same hardware. Ollama produced a response at a rate of 6.62 tokens/sec while Jan AI produced a response at around 5 tokens/sec.

On the second test, Ollama reported 7.35 tokens per second while Jan AI responded at 3 tokens per second. Also, I tried Llama.cpp for the same Gemma 5 E2B model with 3.97 tokens/sec.

The secondary issue is Jan application often freezes your system. I agree, running LLM models takes significant RAM memory, but there are a couple of attempts I faced where giving a prompt with the Gemma 4 E2B model freezes my system. I haven't faced any similar issue with Ollama. That's why I kept Ollama and Llama.cpp for my local AI model inference.

Who is Jan AI actually for?

After spending time with it, I'd say Jan AI fits a few kinds of people particularly well. One thing I need to justify here, although in my tests, Jan struggled to give me more than 5 tokens per second in CPU mode, running it on a system with modern GPUs like Radeon GPU, Intel Arc, Nvidia 4050, you may get to see a huge boost in performance, especially with tokens per second. You could achieve around 13-15 tokens per second.

Privacy-conscious users who want AI assistance without feeding their conversations to a corporation. Lawyers, journalists, healthcare workers, and anyone handling sensitive information fall into this category.

Developers who want a local AI backend for their tools. The OpenAI-compatible API makes it drop-in compatible with a surprising number of existing integrations.

Linux enthusiasts who simply prefer FOSS software and want to own their stack. The fact that it's AGPL-licensed and actively developed is a big plus.

People on a budget. It comes with no API costs or subscription. You run it on hardware you already own.

It's not for everyone. If you need GPT-4 level capability for complex reasoning tasks, you'll still feel the quality gap compared to frontier cloud models. But for writing assistance, summarization, brainstorming, and everyday tasks? A well-tuned 7B or 13B model gets the job done.

The bigger picture

Local AI is no longer a hobbyist experiment. The AI models like Gemma 4, Gemma 3, Granite, Jan AI model, and SmolLM2 have gotten small enough that a standard laptop can run something genuinely useful. Tools like Jan have made the setup approachable enough that you don't need to be a machine learning researcher to get started.

Jan AI won't replace cloud services for everyone. But for a growing number of Linux users who care about privacy, cost, and control, it's become a daily-use tool. I'd encourage you to try it. Download it, grab a 7B model, and see how far it gets you.

Flathub erlässt strengere Regeln für die Einreichung von Apps, bei denen KI zum Einsatz kommt. Aber auch deren Umfeld wie Metadaten und sogar der Pull-Request-Text fallen unter das Verbot.

You’ll have to sift through fewer vibe-coded apps on Flathub in future, as the store has announced a policy change on software made using AI tools. Flathub, the de-facto place to find and install Flatpak applications, is banning the use of “AI” coded applications and automated submissions going forward. It’s not a blanket ban – mature projects with AI code are allowed A change to the store’s policy note says “applications containing AI-generated or AI-assisted code, documentation, or other content are not allowed”. A carve out will allow “mature, well-maintained projects” to include AI generated code and use AI tools […]

You're reading Flathub bans AI-coded apps – with some exceptions, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

Daily builds of Ubuntu 26.10 ‘Stonking Stingray’ are now available for download, as development on the distro’s next major release kicks in to gear. As the name suggests, new ISOs are produced from development code on a (mostly) daily basis, giving those keen to test October’s release in advance the ability to do so. However, because package updates can break the ability for a bootable image to be created, it’s not unusual for there to be temporary gaps between new daily builds being available. Daily builds will continue to be produced for remainder of the Ubuntu 26.10 development cycle, right […]

You're reading Ubuntu 26.10 daily builds now available to download, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

AI features are coming to Ubuntu in 2026, though Canonical has made clear that the distro is not becoming an AI product. In a community post, Jon Seager, VP of engineering at Canonical, says the company is “ramping up its use of AI tools in a focused and principled manner” this year, with a bias toward local inference and open-weight models whose licence terms match Canonical’s values. AI features in Ubuntu will take one of two forms. Implicit features improve existing capabilities using on-device AI models, for things like text-to-speech and speech-to-text to bolster accessibility. Ubuntu will become a context-aware […]

You're reading Canonical is ‘ramping up’ AI in Ubuntu this year, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

Bugs und Sicherheitslücken in Software werden heutzutage vorwiegend von KI gejagt. cURL-Entwickler Daniel Stenberg sieht in den vergangenen Monaten eine Zunahme der Qualität der Einreichungen.

Tails 7.7 erinnert die Anwender an abgelaufene Secure-Boot-Zertifikate und behebt einen Fehler, der das Root-Verzeichnis ausschließlich für Root lesbar

Mozilla baut zum Leidwesen vieler Anwender vermehrt KI-Funktionen in Firefox ein. Jetzt berichtet das Mozilla-Blog, Anthropics neues KI-Modell Mythos habe 271 Sicherheitslücken in Firefox 150 entdeckt und behoben.

Mozilla Thunderbird 150 open-source email client is now available for download with support for generating unobtrusive signatures for OpenGPG, Account Hub improvements, and many other changes.

The post Mozilla Thunderbird 150 Brings Support for Generating Unobtrusive Signatures appeared first on 9to5Linux - do not reproduce this article without permission. This RSS feed is intended for readers, not scrapers.

Your mouth can (probably) say things quicker than your hands can type them, yet voice typing is rarely used as a primary input method on desktop, despite most of us thinking nothing of using it on mobile. That’s despite speech-to-text being available on desktop OSes for decades, natively and through dedicated apps. It never caught on because it was inaccurate and slow and typically hidden away as an assistive feature. (And because a lot of what you do at a keyboard is navigation and that is less efficient to speak, unless ‘arrow down, arrow down, arrow down’ is some trendy […]

You're reading Type with your voice on Linux using this Whisper-based app, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

Thunderbolt is a new open source AI client from the Mozilla-owned MZLA Technologies aimed at enterprises who want to run self-hosted chatbots on their own infrastructure. MZLA Technologies is the for-profit subsidiary of the Mozilla Foundation that develops and maintains the Thunderbird email client. It says Thunderbolt was created with the support of a grant from Mozilla. Terrible name aside (Intel owns a trademark for ‘Thunderbolt’ which Apple markets heavily, so it’s not the best choice for clarity), the LLM that MZLA Technologies asked to write their press release says that Thunderbolt is a “sovereign AI client” for organisations who want their own AI infrastructure. […]

You're reading Thunderbolt is an open-source ‘AI client’ from Mozilla’s for-profit arm, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

Tails 7.6.1 ist ein außerplanmäßiges Notfall-Release, das mehrere Sicherheitslücken aus Firefox ESR 140.9.1 schließt. Ein zeitnahes Update auf die neue Version ist ratsam.

Several Intel NPU fixes land in the latest update to Resources, which Ubuntu is making the default system monitor in 26.04 LTS. Resources is a GTK4/libadwaita tool that shows more system usage, processes and hardware details than GNOME System Monitor, which it is replacing as Ubuntu’s default in the new LTS. The Resources v1.10 release at the start of February added (among other changes) support for AMD Neural Processing Units (NPUs). The app has supported usage for Intel NPUs since its v1.7.x release in late 2024. The v1.10.2 changelog resolves what is described as “implausible” usage values in Intel NPU […]

You're reading Resources 1.10.2 fixes ‘implausible’ Intel NPU values, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

Pigeon Email Notifier is a GNOME Shell extension that does one thing: show a desktop notification when new mail arrives in your Gmail, Microsoft Outlook or IMAP webmail account. If you don’t want to leave a webmail tab open in Firefox, the overhead of a desktop email app like Thunderbird, or your provider doesn’t offer a desktop Linux app (like Fastmail and Proton Mail now do), Pigeon provides a set-and-forget way to still get new mail alerts. Desktop email notifiers have been around for a long time. I’ve written about many standalone tools, like Unity Mail, Popper and Mail Nag over […]

You're reading Get Gmail alerts on Ubuntu without a dedicated mail client, a blog post from OMG! Ubuntu. Do not reproduce elsewhere without permission.

Ein bezeichnender Unterschied zwischen proprietärer und freier Software ist, dass, wenn mit Software Geld verdient wird, die Bedenken schnell wegfallen, während bei Freier Software noch diskutiert wird.

Die auf Pen-Testing und ethisches Hacking spezialisierte Distribution Kali Linux bietet seit Kurzem optional die Verwendung von Claude AI über einen eigenen MPC-Server an.