Sakhanda Wire
NVDA $150.96 -2.89% MSFT $508.29 +0.40% GOOGL $282.44 +0.80% META $616.38 -0.40% AMZN $267.74 +3.82%

Latest news

Older stories

Ars Technica

Nvidia agrees to buy Hugging Face for $13 billion

Nvidia has agreed to acquire Hugging Face, an AI model platform, for $13 billion. The deal is the latest move by the $5.4 trillion chipmaker to use its financial resources to accelerate the expansion of artificial intelligence while gaining greater influence over the industry. Nvidia says Hugging Face will remain open after the transaction.

Hugging Face

NeoMME introduces compact multilingual multimodal encoders

Hugging Face has introduced NeoMME, a family of multilingual multimodal encoders with 260M and 800M parameters. Unlike architectures that combine a pretrained vision tower with a causal language model, NeoMME uses one bidirectional Transformer for text tokens and raw 32×32 image patches. The models were trained from scratch with a masked discrete-diffusion objective and support text, images, code, mathematics and document pages.

TechCrunch

Nvidia to acquire Hugging Face for $12.93 billion

Nvidia has confirmed an agreement to acquire Hugging Face for $12.93 billion, following weeks of reports about a possible deal. Hugging Face’s platform includes three million models, one million applications and half a million datasets, and is used by more than 18 million developers.

The Verge

Nvidia agrees to buy Hugging Face for $12.93 billion

Nvidia has agreed to acquire Hugging Face for $12.93 billion, bringing the open-source AI platform under the ownership of the world’s biggest AI chipmaker. Founded in 2016, Hugging Face lets developers share machine-learning models, datasets, and tools, and is often compared with GitHub because of its searchable repository and community features.

The Decoder

Sam Altman warns AI infrastructure expansion is becoming unsustainable

OpenAI CEO Sam Altman has described the global expansion of AI computing infrastructure as “unsustainable silliness.” In an interview, he said some neocloud providers were announcing very large amounts of capacity without enough customers or revenue to support those plans. Altman contrasted this with OpenAI’s own expansion, which he said is profitable and supported by actual demand.

The Decoder

Meta releases Muse Spark 1.3 with stronger agentic performance

Meta has released Muse Spark 1.3 through Muse Code and the Meta Model API, making it the fourth model in the series introduced within five months. The xhigh tier is available now. The more compute-intensive max tier remains a limited preview for selected partners while additional safety testing continues. Meta also says an open-weights version is planned.

The Decoder

Anthropic signs $35 billion cloud deal with Lambda

Anthropic has signed a $35 billion cloud-computing agreement with Lambda, a cloud provider backed by Nvidia, according to Reuters, which cited an anonymous source. The planned data center is in Nueces County, Texas, and is expected to provide about 350 megawatts of capacity.

MarkTechPost

Perplexity Open-Sources Lily Inference Engine for Apple Silicon

Perplexity has released Lily as open source, presenting it as the local inference engine used by Hybrid Compute in Perplexity Computer. The runtime is written in Rust and uses custom Metal kernels, without PyTorch or MLX in its execution path. A public demo in the pplx-garden repository provides greedy generation through a minimal OpenAI-compatible HTTP API.

Hugging Face

GRPO fine-tuning lifts LFM2.5-350M structured-output score

A public training recipe uses Group Relative Policy Optimization (GRPO) to improve structured-output compliance in LiquidAI’s LFM2.5-350M model. The procedure relies on the TRL library, about 500 training samples and 100 steps, allowing the run to fit a free-tier Colab or Kaggle GPU. Evaluation uses the open IFStruct benchmark, which checks whether responses are valid and conform to requested schemas and structures.

Hugging Face

funes Gives Coding Agents Persistent, Local Memory

Hugging Face has introduced funes, an open-source memory layer for coding agents including Claude Code, Codex, pi and Hermes. The tool turns existing agent session traces into a searchable local memory, preserving decisions, failed approaches and their rationale instead of leaving them in separate logs. A single command creates an index, adds recall and retrieval tools, and updates the index after completed turns.

Hugging Face

Open pipeline trains a coding model to paint watercolours

An open implementation explores whether reinforcement learning can teach a language model to generate watercolour-style paintings. The project, described by Sergio Paniego, follows Surya Narreddi’s earlier work, in which a model writes JavaScript using p5.brush, a library that simulates features such as pigment bleeding, paper texture and brush flow. Narreddi’s video had received more than 1.5M views when the article was written.

MarkTechPost

Qwen Developers open-source local search layer zg

Qwen Developers has released zg, also called zvec-grep, an open-source search layer for people and coding agents. Licensed under Apache 2.0, it combines ripgrep, BM25 keyword ranking and vector search behind one interface. The package is available through npm as @zvec/zvec-grep and requires Node.js 22 or newer on macOS, Linux or Windows. Its default model works without a GPU.