Search: “model-releases”

Meet Qwen-Image-2.1, the most balanced and cost-effective image generation model in the Qwen-Image series! Now open weights! 🎨 A unified model for both generation and editing, delivering top-tier qual

Released: Update your ComfyUI for workflow Weights: https://huggingface.co/Comfy-Org/Qwen-Image-2.1/tree/main Variety of Examples https://youtube.com/watch?v=tP_gGGJzEqg&feature=youtu.be https://model

Qwen‑Image‑2.1, released by Alibaba’s Qwen team, is an image generation model that handles a broad spectrum of editing tasks while maintaining balanced performance across them. For instance, when prov

Overclocking the CMP 170HX 40GB I was able to get the memory bandwidth from 1,386.2 GB/s to 1,890.1 GB/s, that's a +36.4% increase. Qwen 3.8 27B token generation jumped from 110 T/S to 202 T/S, same c

​ ​I wanted to share a quick update and performance video running the full Moonshot AI Kimi K3 (moonshotai/Kimi-K3) model across my 16x GB10 cluster. ​Getting a 2.8T parameter model running smoothly r

Sharing my successful experience here. My setup: RTX 5060 Ti 16GB (x16 PCIe slot) RTX 3060 12GB (x4 PCIe slot) RAM: 32GB CPU: Ryzen 5 5600 OS: Linux So I have a total of 28GB. Average tk/s: 50, and th

Thanks @vllm_project for the day-0 support! 🙌 Generate, edit, transparent output — one model, ready to serve. Details 👇 Qwen-Image-2.1 from @Alibaba_Qwen has day-0 support in vLLM-Omni. 🎉 Generate, ed

TL;DR: Local agent loop, ~21 days, one RTX 3090. Task was pretty much 'build a CUDA inference engine for optimized for yourself on this GPU arch.' Got working kernels and benches, not a win over llama

I have been anti-low quant. I can feel the difference, I swear. I am also anti-quantized KV cache. I have been burned there - I think the tradeoffs compound, especially at longer contexts. Due to this

Josipa Majic Predin / Forbes: Vercel, Cloudflare, and others quickly add Jev, as it makes AI tool selection much faster and cheaper; TypeSafe: Jev matches GPT-5.6 and Sonnet 5 workflow evals — Most of

hey! we're a small swiss and south african lab and we just put out hemmingway-1. it's a 27b model we trained specifically for roleplay, creative writing, dialogue, stories, and for everyday texting. f

We're excited to announce our AI Engineer New York speaker lineup! Presented by @arizeai CEO @ExaAILabs CTO @tryramp CEO @ValsAI CEO @RogoAI CTO @modal CEO @trybasis CEO @SonarSource CEO @runwayml CTO

I am trying to learn more about long-term memory and how people are handling that nowadays. Historically, I have used a mixture of paid cloud AI services (ChatGPT, Gemini, OpenRouter API) and my own l

First time trying something like this. I asked chatgpt and it told me there are mcp server plugins for both ghidra and ida pro as well as RE-MCP which has support for both. I could run that on my low

In February OpenAI stopped reporting SWE-bench Verified and recommended other labs stop too. Every frontier model they tested could reproduce the human-written reference fix, or verbatim details of th

Just find it interesting, since if Taalas tried to get into the consumer market, it could be interesting. Kind of like how we have game discs on CDs. Of course, a Qwen3.8-27B chip is stuck on 3.8-27B,

Yes, you can make up some of the difference by giving Claude API access but (1) most people will never do this, (2) it means additional expense or difficulty, and (3) the real gain isn't from image ge

Behold, workflow whisperers! YuE2 was released a week ago and I gotta say it's splendid for its size (3B). My favourite local music generation model at the moment, no doubt. The good news is, I've fou

A great example of practical AI solving real-world problems! Faster than ever. Thanks for building with Qwen. @cerebras Try it out for yourself! 🏠 We built Money Agent, a personal finance assistant po

A new release of our Agent Skills and Docs eval library, Cultivar is available! Now, use @typesafeai's Jev model to evaluate how well your agents use docs and Skill against specified tasks in Modal sa

Anthropic, an AI safety and research organization, has established a “wet lab” in the Bay Area to enable the Claude language model to perform real‑world biology experiments. The move accompanies effor

cuda : fix CUB argsort corruption caused by in-place keys (#28389) argsort_f32_i32_cuda_cub called the one-shot DeviceRadixSort::SortPairs API with d_keys_in == d_keys_out (temp_keys, temp_keys). CUB'

metal : support qwen4exp hc ops (#29000) Add support for the new DSV4 HC op variants used by qwen4exp: hc_pre with per-element sigmoid gate (gated variant) hc_post with identity mixing (comb == nullpt

test-llama-archs : generate dummy test vocab (#29084) Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations

metal : fix FA support checks (#29122) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/48626162 macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (

json-schema : accept escaped hyphen in regex patterns (#29127) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/48635641 macOS/iOS: macOS Apple Silicon (arm6

server : improve startup log messages (#29125) server-models : show source per model in log Show [source] tag (preset/models_dir/cache) per model instead of cryptic * marker Show HF hub cache path in

hexagon: enable support for TOP_K op (#29113) hexagon: enable support for TOP_K op hex-topk: thread single-row TOP_K, raise VTCM-based size cap hex-topk: fix TOP_K mdev row partitioning hex-topk: opti

hexagon: add support for GEGLU_QUICK (#29114) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/48660160 macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Si

hexagon: enable I32 GET_ROWS (#29116) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/48662388 macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (a

Banger paper from Google Cloud AI Research. If you follow autonomous research agents, this one is worth your time. ScientistTwo takes a problem from a human expert and runs the full discovery cycle wi

Banger paper from MIT and Sakana AI. They show that self-improving coding agents work. The best part is that their approach, Self-Improvement via Fast Tree-search (SIFT), runs at a tenth of the CPU ho

I needed a relatively simple but acceptable level of AI for working on one project. I didn't have any heavy requests, I just needed to give the AI access to the project files so it could search throug

I’ve used both ChatGPT and Gemini, and lately I’m consistently getting better results from ChatGPT — especially with prompt following, realism, composition, and text. The improvement over time is hone

Release: datasette-auth-github 1.0 I run this GitHub login plugin on the agent.datasette.io demo site and I noticed that my authenticated sessions weren't lasting very long. It turned out that the plu

Play it right in your browser, no download. Some of the GLB’s are messy, but overall I’ve enjoyed playing with it, and the last iteration made was the gun/weapon placement. Contributors are welcome, w

Although I do not completely understand what the big deal is with Jev, it reminded me of something I implemented in mt_llm (which is build around the great llama.cpp) a while back to use a relatively

Hey everyone! I want to share the results of a small language model architecture experiment I’ve been running as a side project. The ultimate goal is to test an approach to reduce the number of the mo

Screenshot of heavy load. I did offload vision to a spare 2060 instead of local ram. The control panel was generated with qwen3.8-27b davidau q4k_m max mtp. Made the panel from scratch over a week, ba

In May, Gemini broke containment and hacked three different companies, but Google didn't disclose the incident until the Wall Street Journal approached the company. The hacks happened during a test of

Allie K. Miller’s tweet (posted Sep 19 2026) states that she believes “Instinct and other similar platforms (Muse, Grokbot, etc.) are better overall experiences than standard ChatGPT or Claude.” The c

Top 3 papers on HF Daily Paper are all unusually delightful and interesting reads for anyone on the leading edge of local LLMs, agent harness optimization, etc, felt like sharing. DeepSeek-V4.1-Flash:

I did a small test of the new hyped quantisation of Qwen3.8 vs the biggest quant which fits into my limited 16GB VRAM with decent context. The results are interesting. Of course, the smaller file give

I think AI automated labs should be banned from anything viral or pathogen related. They will do gain of function research otherwise. Anyone agree with me? BREAKING: Anthropic has quietly set up a 'we

TL;DR: I run a mismatched Tesla V100-PCIE pair—one 16 GB card and one 32 GB card, 48 GB total—in a Proxmox/LXC-based local-inference lab. The practical winner so far is a recent CUDA build of llama.cp

I improved the TPS of Gemma 4 31B. Improving TPS and performing optimisations requires understanding of the model architecture, and I had to fork VLLM and apply my own patch to break into making a con