Local AI - Page 8
Stories from Local AI come first, newest first, then stories from connected topics (llama.cpp, Ollama, NVIDIA RTX Spark, on-device AI, AI Models, Hugging Face), each labelled with where it comes from. This view is not indexed by search engines.
Latest
fal Launches FLUX 3 Max Video Model, Built With Black Forest Labs
fal and Black Forest Labs launch FLUX 3 Max, a video model fal says makes a 10-second clip in under 11 seconds, with a preliminary 1,139 Elo.
OpenAI Says Its Latest Models Are Far Better at COBOL and Fortran Code
OpenAI says GPT-6 Astra and GPT-6.1 Sol score far higher on its own COBOL and Fortran coding evals, though production use still needs expert review.
Amazon Bedrock Brings SpaceXAI’s Grok 4.7 to AWS GovCloud (US)
Amazon Bedrock now offers SpaceXAI's Grok 4.7 in AWS GovCloud (US), AWS says, bringing the coding and agent model to US public-sector customers.
Together AI Lets Teams Write Their Own RL Loops on Open Models
Together AI launches Together Post-Training for custom RL and SFT on open models, at $9 per B200 GPU-hour for eight weeks. Customer quotes are supplied.
Indonesia Gets a 27B Open-Weight Model Tuned for Local Languages
GoTo’s 27B Gemma-based model covers Indonesian plus five regional languages. Open-weight under Google’s Gemma licence; AI Singapore co-developed.
NVIDIA Releases Nemotron Labs Dory1.0-4B Looped Open-Weight Model
NVIDIA’s Nemotron Labs Dory1.0-4B is a 4B open-weight looped model: one checkpoint, 1–4 loops per token. OpenMDW-1.1; SWE-bench figures are NVIDIA’s.
Reflection Releases Beam Open Weights on Hugging Face Under Apache 2.0
Reflection has published Beam's weights on Hugging Face: a 501B mixture-of-experts model, 23B active, with a 1M-token context and an Apache 2.0 licence.
FPT and NVIDIA Expand Nemotron-Personas-Vietnam Dataset to 1 Million Records
FPT and NVIDIA expanded Nemotron-Personas-Vietnam from 100,000 to 1 million synthetic profiles. Profiles are AI-generated; Hugging Face lists CC BY 4.0.
IBM Releases Granite Switch 4.1 With Decision Adapters and Fast Speech Recognition
IBM’s Granite Switch 4.1 packs Decision and Guardian aLoRAs plus Speech-5.0-Turbo-CTC in one endpoint. Leaderboard and speed figures are IBM’s claims.
NVIDIA and AI4Bharat Detail a 2.8-Trillion-Token Indic Pretraining Corpus, Releasing Soon
NVIDIA and AI4Bharat detail a ~2.8T-token Indic corpus for 22 languages, releasing soon, plus an open GPU pipeline. Companion to Nemotron 3.5 Super VL.
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.