All posts tagged: opensource

I stopped using LM Studio once I found this open-source alternative

I stopped using LM Studio once I found this open-source alternative

Published by skeptic

LM Studio has been my go-to app for running local LLMs since I discovered it. It’s easy to use, the UI is well-polished, and you can have any AI model up and running in a matter of a few clicks — even if you’ve never touched a terminal in your life. It’s one of the best tools if you want to enjoy the benefits of a local LLM. But while LM Studio is free to use, it isn’t open-source. Some of its components, especially the command-line tooling, carry open licenses, but the app itself is proprietary software. For a tool that can easily sit at the core of your local AI workflows, risking a license change whenever the parent companies want is a risk I was uncomfortable with, and then I found Jan. Related I’ll never pay for AI again AI doesn’t have to cost you a dime—local models are fast, private, and finally worth switching to. I thought I’d miss LM Studio. I didn’t Why Jan feels like a real replacement for LM Studio …

This open-source app made every download manager I used before feel unnecessary

This open-source app made every download manager I used before feel unnecessary

Published by skeptic

I have a complicated relationship with download managers. While some purists argue that the Command Prompt is actually the best download manager for Windows, I’ve always sworn by Internet Download Manager, which, in itself, is considered one of the best file download managers out there.I tolerated its garish toolbar and the annual license renewal that always seemed to sneak up on me right when I’d forgotten it existed. Things got messier during a stretch when I was splitting time between Linux and trying to find a capable Linux download manager. Then the cracks really showed: one app for torrents, a separate client for magnet links, and the browser fumbling through everything else with inconsistent results. I had four tools doing the job of one. Then a Reddit discussion thread mentioned Gopeed in passing — specifically the phrase “it just handles everything.” I clicked mostly out of habit, expecting another niche downloader I’d forget about in a day or two. That was several weeks ago, and I haven’t gone back to my old setup since. Now, …

Google’s Gemma 4 model goes fully open-source and unlocks powerful local AI – even on phones

Google’s Gemma 4 model goes fully open-source and unlocks powerful local AI – even on phones

Published by skeptic

Elyse Betters Picaro / ZDNET Follow ZDNET: Add us as a preferred source on Google. ZDNET’s key takeaways Gemma 4 is now fully open-source under Apache 2.0. Local AI enables privacy, offline use, and lower costs. From servers to smartphones, deployment just got much easier. Google announced today that its DeepMind AI research division is releasing Gemma 4, its latest generation of open large language models. The models are being released under the Apache 2.0 license, making them truly open source compared to the permissive but still controlled license of earlier Gemma generations. What is Gemma? Gemma is an LLM like Gemini. But here, we’re talking about the AI processing engine, not the chatbot interface. Both Gemma and Gemini were developed using the same research and technology. The difference is that Gemini is a subscription-based closed product, whereas Gemma is an open model that can be downloaded and run locally for free. The ability to run an AI model locally without a fee benefits a variety of applications. There are plenty of folks who want to run …

Mercor says it was hit by cyberattack tied to compromise of open-source LiteLLM project

Mercor says it was hit by cyberattack tied to compromise of open-source LiteLLM project

Published by skeptic

Mercor, a popular AI recruiting startup, has confirmed a security incident linked to a supply chain attack involving the open-source project LiteLLM. The AI startup told TechCrunch on Tuesday that it was “one of thousands of companies” affected by a recent compromise of LiteLLM’s project, which was linked to a hacking group called TeamPCP. Confirmation of the incident comes as extortion hacking group Lapsus$ claimed it had targeted Mercor and gained access to its data. It’s not immediately clear how the Lapsus$ gang obtained the stolen data from Mercor as part of TeamPCP’s cyberattack. Founded in 2023, Mercor works with companies including OpenAI and Anthropic to train AI models by contracting specialized domain experts such as scientists, doctors, and lawyers from markets including India. The startup says it facilitates more than $2 million in daily payouts and was valued at $10 billion following a $350 million Series C round led by Felicis Ventures in October 2025. Mercor spokesperson Heidi Hagberg confirmed to TechCrunch that the company had “moved promptly” to contain and remediate the security …

How AI has suddenly become much more useful to open-source developers

How AI has suddenly become much more useful to open-source developers

Published by skeptic

imaginima/ iStock / Getty Images Plus via Getty Images Follow ZDNET: Add us as a preferred source on Google. ZDNET’s key takeaways Top open-source maintainers find that AI has suddenly become much more useful. There are still legal and ‘AI slop’ problems to overcome. By year’s end, AI programming tools should be much more reliable. With open-source software running pretty much everything, you might think that multiple developers maintain most of the important programs with help from corporate sponsors. You’d be wrong. As Josh Bressers, VP of security at software supply-chain company Anchore, pointed out last year, the vast majority of open-source projects, 7 million out of 11.8 million programs, have only a single maintainer. You might think that those programs are obscure or no longer used. You’d be wrong about that, too.  Also: 7 AI coding techniques I use to ship real, reliable products – fast Bressers looked closely at the JavaScript NPM ecosystem and found that, among the projects downloaded over a million times a month, “about half of the 13,000 most downloaded NPM packages are [maintained …

Windows’ default file copy is embarrassingly slow compared to this free open-source alternative

Windows’ default file copy is embarrassingly slow compared to this free open-source alternative

Published by skeptic

Windows Explorer does basic file operations just fine. Drag and drop a few files, and they move in an instant. However, if you are moving a large folder with mixed file types, it chokes and slows to a crawl. This comes from a design choice in how Windows stages file operations. While you can use a third-party app like FastCopy to move large files faster, Windows has always had RoboCopy, a built-in command-line utility that handles large transfers significantly better. But it’s less popular due to its command-line interface. ChoEazyCopy wraps RoboCopy in a proper GUI, making it accessible without you needing to memorize a single command. What’s RoboCopy, and why is it faster than File Explorer? A powerful built-in copy tool hidden behind the command line There’s a simple reason why File Explorer takes forever to start copying files. It stages every operation before transferring, processing files one at a time. This approach works for a handful of files, but with thousands of small files or deeply nested folders, the progress bar slows to …

Nvidia’s Nemotron-Cascade 2 wins math and coding gold medals with 3B active parameters — and its post-training recipe is now open-source

Nvidia’s Nemotron-Cascade 2 wins math and coding gold medals with 3B active parameters — and its post-training recipe is now open-source

Published by skeptic

The prevailing assumption in AI development has been straightforward: larger models trained on more data produce better results. Nvidia’s latest release directly challenges that size assumption — and the training recipe behind it may matter more to enterprise AI teams than the model itself. The open-weight model’s Cascade RL post-training pipeline, detailed in Nvidia’s technical report, offers a reproducible blueprint for enterprise teams building domain-specific reasoning systems without training from scratch. Nemotron-Cascade 2 is an open-weight 30B Mixture-of-Experts (MoE) model that activates only 3B parameters at inference time. Despite this compact footprint, it achieved gold medal-level performance on three of the world’s most demanding competitions: the 2025 International Mathematical Olympiad (IMO), the International Olympiad in Informatics (IOI), and the ICPC World Finals. It is the second open model to reach this tier, after DeepSeek-V3.2-Speciale — a model with 20 times more parameters. Why post-training is becoming the real competitive advantage Pre-training a large language model from scratch is enormously expensive — on the order of tens to possibly hundreds of millions of dollars for frontier …

I found a browser-based creative suite and it being open-source is only part of the appeal

I found a browser-based creative suite and it being open-source is only part of the appeal

Published by skeptic

I have a love-hate relationship with creative software. The tools that are good enough to rely on professionally can cost a small fortune, and the free alternatives come with their own learning curve and drawbacks. That said, the fact that open-source alternatives for Adobe programs exist gives me hope. That hope led me to Graphite, a browser-based, fully open-source 2D design tool. I tried it, thinking it’d be like another browser-based Photoshop clone. Instead, I found a fully capable design tool I can spin up in any browser, any machine. The best part? It has very little to do with the price tag. It runs in your browser, but doesn’t act like it Yes, Graphite runs in your browser, whatever that may be. It’s free to use, and the source code is available in the official GitHub repository under an Apache 2.0 license. But considering it as just another free Illustrator or Photoshop alternative you can open in a tab massively undersells what it’s meant for. There are creative tools that are free, and professionals …

I was losing track of my self-hosted apps until I found this open-source dashboard

I was losing track of my self-hosted apps until I found this open-source dashboard

Published by skeptic

One of the first signs that things were getting out of hand was the day I needed to figure out which apps were running, and after opening Portainer, I still guessed the wrong port. At the time, I had self-hosted tools on my NAS and a small server. I had to remember the port numbers for each app or bookmark them, but I just wasn’t able to properly keep track. Even though my tools were not failing, the real issue was that I had so much difficulty finding what I’d already set up. I stumbled upon DockPeek while trying to fix a reverse proxy issue, and as soon as I installed and opened this open-source tool, it was clear I had found the solution I desperately needed. OS Windows, MacOS, Linux Price model Free Dockpeek is a lightweight, self-hosted Docker dashboard. It allows quick access to your containers and lets you open web interfaces, view logs, and monitor your ports. I just couldn’t see my setup Every app turned into a port number I had …

z.ai debuts faster, cheaper GLM-5 Turbo model for agents and ‘claws’ — but it’s not open-source

z.ai debuts faster, cheaper GLM-5 Turbo model for agents and ‘claws’ — but it’s not open-source

Published by skeptic

Chinese AI startup Z.ai, known for its powerful, open source GLM family of large language models (LLMs), has introduced GLM-5-Turbo, a new, proprietary variant of its open source GLM-5 model aimed at agent-driven workflows, with the company positioning it as a faster model tuned for OpenClaw-style tasks such as tool use, long-chain execution and persistent automation. It’s available now through Z.ai’s application programming interface (API) on third-party provider OpenRouter with roughly a 202.8K-token context window, 131.1K max output, and listed pricing of $0.96 per million input tokens and $3.20 per million output tokens. That makes it about $0.04 cheaper per total input and output cost (at 1 million tokens) than its predecessor, according to our calculations. Model Input Output Total Cost Source Grok 4.1 Fast $0.20 $0.50 $0.70 xAI Gemini 3 Flash $0.50 $3.00 $3.50 Google Kimi-K2.5 $0.60 $3.00 $3.60 Moonshot GLM-5-Turbo $0.96 $3.20 $4.16 OpenRouter GLM-5 $1.00 $3.20 $4.20 Z.ai Claude Haiku 4.5 $1.00 $5.00 $6.00 Anthropic Qwen3-Max $1.20 $6.00 $7.20 Alibaba Cloud Gemini 3 Pro $2.00 $12.00 $14.00 Google GPT-5.2 $1.75 $14.00 $15.75 …