Skip to content
Trending storyDeveloping

Ant Group open-sources Ling-3.1-flash model weights

1 article1 sourcesince Oct 14Last article 5h ago ·

Overview

AISummary of 1 article

Ant Group's AntLingAGI releases Ling-3.1-flash as open weights, with about 560B total parameters, about 25B active per token, and up to 1M context.

The release is reported by AntLingAGI's announcement, as relayed in an X post.

A separate post from Atomic Chat says Ling 3.1 Flash, run locally, beat GPT-6 Luna on a Three.js LEGO Eiffel Tower prompt at 2.8x lower cost. That comparison is the post's own claim and has not been independently verified.

Written by AI from the articles below · updated Oct 14, 11:23 PM ET

Check the sources:

Latest developmentOct 14, 7:16 PM ET
Ling 3.1 Flash open weights beat GPT-6 Luna on a LEGO prompt

Article timeline

The articles in this story. Times are ET.

Oct 14
  1. atomic.chatX
    Ling 3.1 Flash open weights beat GPT-6 Luna on a LEGO prompt

    AIAnt Ling released Ling-3.1-flash open weights, with about 560B total parameters, about 25B active per token and up to 1M context. A Atomic Chat post says Ling 3.1 Flash, run locally, beat GPT-6 Luna on the same Three.js LEGO Eiffel Tower prompt at 2.8x lower cost.

    Video from @atomic_chat_hq's post

Heat trend

Not enough continuous observations to show a trend yet.