13+ hour, 23+ min ago  (766+ words) NVIDIA Nemotron-3-Nano-Omni-30 B-A3 B is an open 30 B parameter, 3 B active hybrid reasoning Mo E model built for multimodal agentic workloads including audio, video, text, images and docs as input, with text output. The model runs on 25 GB RAM for…...

1+ week, 1+ day ago  (1031+ words) Run the new Qwen3. 6-35-A3 B model locally! Qwen3. 6 is Alibaba's new family of multimodal hybrid-thinking models, including Qwen3. 6-35 B-A3 B. It delivers top performance for its size, supports 256 K context across 201 languages. It excels in agentic coding, vision, chat tasks. 35 B-A3 B GGUFarrow-up-right can run on…...

1+ week, 2+ day ago  (541+ words) Use open models with Open AI Codex on your device locally. This guide will walk you through connecting open LLMs to the Codex CLI entirely locally. It works with any Open AI or API compatible local model setup including: Deep…...

2+ week, 3+ day ago  (769+ words) Run Mini Max-M2. 7 LLM locally on your own device! Mini Max-M2. 7 is a new open model for agentic coding and chat use-cases. The model achieves SOTA performance in SWE-Pro (56. 22%) and Terminal Bench 2 (57. 0%). The 230 B parameters (10 B active) model has a…...

2+ week, 3+ day ago  (1260+ words) Unsloth Changelog for our latest releases, improvements and fixes. To use the latest changes, update Unsloth via unsloth studio update. We've updated Gemma 4 with many fixes. These bugs are universal and affected all training packages and implementations and did not…...

3+ week, 22+ hour ago  (742+ words) Run the new GLM-5 model by Z. ai on your own local device! GLM-5 is Z. ai's latest reasoning model, delivering stronger coding, agent, and chat performance than GLM-4. 7, and is designed for long context reasoning. It increases performance on benchmarks such…...

3+ week, 14+ hour ago  (844+ words) Run the new GLM-5. 1 model by Z. ai on your own local device! GLM-5. 1 is Z. ai's new open model. Compared with GLM-5, it delivers major improvements in coding, agentic tool use, reasoning, role-play, long-horizon agentic tasks, and overall chat quality. The…...

3+ week, 5+ day ago  (997+ words) Run Google's new Gemma 4 models locally, including E2 B, E4 B, 26 B A4 B, and 31 B. Gemma 4 is Google Deep Mind's new family of open models, including E2 B, E4 B, 26 B-A4 B, and 31 B. These multimodal, hybrid-thinking models support 140+ languages, up to 256 K context, and come in…...

3+ week, 5+ day ago  (717+ words) Train Gemma 4 by Google with Unsloth. You can now fine-tune Google's Gemma 4 E2 B, E4 B, 26 B-A4 B and 31 B with Unslotharrow-up-right. Support includes all vision, text, audio and RL fine-tuning. Fine-tune Gemma 4 via our free Google Colab notebooks: If you want to…...

1+ mon, 1+ week ago  (324+ words) A guide for getting started with the fine-tuning studio, data recipes, model exporting, and chat. Unsloth Studio is a local, browser-based GUI for fine-tuning LLMs without writing any code. It wraps the training pipeline in a clean interface that handles…...