60 بالعربي
HomeWatchListen
Breaking
US Inflation Reaches 3.7% and Airline Fares Surge 22.4% Annually•Egyptian Housing Ministry Announces 6 Lotteries and EGP 10 Billion Financing to Acquire Assets•China Proposes Draft Export Controls on 40 Technologies, Including 23 in AI•US Inflation Reaches 3.7% and Airline Fares Surge 22.4% Annually•Egyptian Housing Ministry Announces 6 Lotteries and EGP 10 Billion Financing to Acquire Assets•China Proposes Draft Export Controls on 40 Technologies, Including 23 in AI•US Inflation Reaches 3.7% and Airline Fares Surge 22.4% Annually•Egyptian Housing Ministry Announces 6 Lotteries and EGP 10 Billion Financing to Acquire Assets•China Proposes Draft Export Controls on 40 Technologies, Including 23 in AI•US Inflation Reaches 3.7% and Airline Fares Surge 22.4% Annually•Egyptian Housing Ministry Announces 6 Lotteries and EGP 10 Billion Financing to Acquire Assets•China Proposes Draft Export Controls on 40 Technologies, Including 23 in AI•

Meta Launches Open Model with 816 Billion Parameters Running on Mac

Meta has launched the open Muse Spark Open model with 816 billion parameters and a context window of one million tokens, stating that it is the most powerful open-weight model in America and can be run locally.

October 17, 2026
Meta Launches Open Model with 816 Billion Parameters Running on Mac

Meta has launched the open-weight "Muse Spark Open" model, stating that it is the most powerful of its kind in the United States, according to TechCrunch.

The model features 816 billion parameters, with a context window of up to one million tokens and native support for multiple data types, and was designed for coding and long-running agent tasks.

The model can be run locally on Apple silicon, and the company demonstrated it on two Mac Studio devices powered by M5 Ultra processors linked via Apple's MLX framework.

The model scored 61.8 on the JobBench benchmark for agents according to Meta's own results, a test used by the company to measure model performance in multi-step tasks.

What do these terms mean?

Parameter: The internal numerical value that the model learns during training; the higher the number, the better the model becomes at representing data.

Context Window: The maximum amount of text that the model can read and reference at one time.

Open-weight: A model whose weight values are published to the public, allowing it to be run or trained locally instead of relying solely on a cloud service.

Share
Keywords

Weekly Newsletter

Read between the lines before everyone else. Decode the most important economic, tech, and decision-maker movements in the region.. in 5 minutes every Saturday.

Follow Us