Open Source AI News - Page 23 of 25
497 articles • Page 23 of 25
Meta's 30B Muse Glimmer AI Model Is Built for Agent Operations
Meta put out a 30-billion-parameter model called Muse Glimmer on Wednesday, built to run autonomous AI agents on a high-end Mac or PC instead of a...
NVIDIA's Groq 3 LPX Enters Full Production for AI Agents
NVIDIA's Groq 3 LPX chip has moved into full production, and the company is pairing it with its Vera Rubin NVL72 rack-scale system to chase a...
NVIDIA RTX 5090 Powers Local AI Agents for Private Files and Apps
NVIDIA is spending August highlighting how far local AI has come, and the timing lines up with a notable release from Meta. On Tuesday, Aug.
New Sparse Embedding Model Scores 58 on BEIR-13 Benchmark
Linkup's research team put out SPARSEUP this week, a learned sparse embedding model built on a 149M-parameter ModernBERT backbone and released under...
Nous Research in Talks for New Funding at USD 1.5 Billion Valuation
Nous Research is closing in on a new funding round that would value the AI startup at $1.5 billion, according to three people familiar with the deal.
How LangExtract Uses URLs and Text Lists for Data Extraction with LLMs
Data extraction tools are usually sold as magic. LangExtract is just plumbing. It's an open-source Python library that tells a large language model...
Hugging Face releases ml‑intern, an agent that auto‑diagnoses LLM failures
Training an LLM is only half the battle. The real drudgery begins after: poring over evaluation logs, spotting subtle regressions, and retuning...
NVIDIA Cosmos 3 supports unified I/O modalities, open‑sources six AI datasets
Physical AI is a mess. It's a pile of disconnected specialists. A model sees images, another controls a robot, a third guesses how a stack of boxes...
Trusted access enables defenders through verification and accountability
OpenAI's blog post on Wednesday ditched the usual lofty pronouncements. It led with cash: a $10 million Cybersecurity Grant Program.
Cursor's Composer 2 built on Chinese AI model, exposing Western open-source gaps
The revelation that Cursor’s Composer 2, a tool designed to wrestle with monster 256,000-token contexts, chose a Chinese foundation model over the...
GPT Open-Source Model Hits 2,988 Tokens/Sec with Low USD 0.45 Per Mil Cost
Open source models have always been cheaper. Now, they’re genuinely fast. Take the GPT OSS 120B: it clocks 2,988 tokens per second.
NVIDIA’s 100M-Parameter Model Tracks 8 Speakers in Real Time
NVIDIA put a new speaker diarization model on Hugging Face this week, and the numbers are small on purpose.
NY RAISE Act targets AI safety as Zhipu AI launches GLM 4.7, Nvidia buys Groq
The AI industry is moving faster than the laws meant to contain it. New York just answered with the RAISE Act, the second major U.S.
OpenAI AI Agents Found Communicating on German Network
A German-language wiki called DseWiki became an unlikely message board this year, not for people, but for AI agents.
White House Slaps Export Controls on Anthropic After 24-Hour Whirlwind
Andrey here, back after a long break, and there's no easy way to ease into this one.
Deepseek Adapts Open-Source AI Software for Huawei's Ascend Chips
Deepseek posted the announcement on its official WeChat channel: the Hangzhou AI lab is releasing open-source programming tools built for Huawei's...
Run Local AI on 8GB Macs With Smaller Models, Avoid Complex Setup
The fantasy of running your own AI isn't about freedom or digital sovereignty. It's about wanting to type a stupid question about lunch without it...
Floqer's AI agents research sales prospects so teams don't have to
Floqer will take the stage Wednesday at TechCrunch Disrupt 2026 to pitch an idea that sounds almost too simple: let AI agents do the grunt work of...
pxpipe hides text in PNGs to cut Claude token costs by up to 70%
Everyone's hunting for cheaper AI bills. Now there's pxpipe. This new open-source tool employs a brutally simple hack: hide your text inside a...
Cognition's SWE-2 Coding Model Matches Rival at 64% Lower Cost
Cognition put a number on its newest coding model Tuesday: 50.0% on FrontierCode 1.1 Main, one point behind Fable 5.1, at 64% of the cost.