Daily AI news, sourcedUpdates / All
Updates 1960 items · newest issue first · page 38 of 98
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- NuMind open-sources a vLLM extension that returns a served LLM's last hidden state as an embedding
- Gewell 0.2 inference engine adds Gemma 4 26B A4B and multimodal EmbeddingGemma 2 servingOfficialGitHub leDissolution/gewellOct 12Models
- Qwen introduces QwenGyre, an elastic RL framework for training agents on extremely long tasks
- BrokenML benchmark tests how much cluster access AI agents need to fix broken ML jobs
- Graphsignal adds graphsignal-top, a live top-style terminal view of inference GPUs and kernels
- White House restores the UK AI Security Institute's early access to Anthropic, Google DeepMind and OpenAI models, per The DecoderReportedThe DecoderOct 12Policy & safety
- Former Plaud exec Peter Mo raises $26M for Ambi, an always-listening AI assistant app, per TechCrunch
- Atticus is an open-source hold-to-talk voice assistant for macOS that runs on Claude Code
- Kaiming He's team releases VISTA, a visual harness that lets multimodal models re-inspect past frames, per QbitAI
- Open Academic Paper Gen drafts papers and checks citations against full texts, per QbitAI
- Zhipu quietly tests GLM-5.3-Highspeed in Zcode, with a user clocking 490 tokens/s, per QbitAI
- Huawei's Memento 3 agent clears all 25 ARC-AGI-3 public games with 56% fewer actions than humans, per QbitAIReportedQbitAIOct 12Research
- TinyDetect runs AI text and image detectors fully in the browser
- CL4SM opens a call for contributions on constrained learning for scientific modelling ahead of NeurIPS Paris
- Linus Torvalds uses Claude Code for a Linux page-cache patch series, per Phoronix
- Matthew Schwartz's BootLoops toolkit lets LLMs carry exact methods into 22 scientific fields
- EXL3-QLoRA brings native QLoRA training to ExLlamaV3 quants at 2-8 bpwUnconfirmedr/LocalLLaMAOct 12Agents & coding
- Open-source DeepSeek Harness plugin analyzes the quality of novel writingUnconfirmedr/LocalLLaMAOct 12Apps & products
- vLLM fork serves Qwen3.8-Flash-Next to 7 users at 262k context on six RTX 3090sUnconfirmedr/LocalLLaMAOct 12Models
- whatmodel.sh ranks the best AI models your own machine can run