LaunchAI ModelsAugust 3, 2026

Ant Group releases Ling 3.0 Flash, a 124B open-weights MoE model

Ling 3.0 Flash is a 124B-parameter MoE with 5.1B active per token, scoring 38 on the Artificial Analysis Intelligence Index. Weights are on Hugging Face under MIT license, with official FP8 and MXFP4 versions. It runs in thinking and non-thinking modes with a 256K context window.

How this story unfolded

4 weeks · 2 reports · 12 community posts · 14 of 15 shown

Jul 23
- Ling 3.0 Flash is now available on AI Gatewayvercel.com
Jul 25
- inclusionAI (Ant Group) released Ling-3.0-flash, a 124B sparse MoE (5.1B active), 256K context, API-only on OpenRouter and free for a week
Jul 27
- Ling-3.0-flash weights: SGLang says day-0, vLLM says when they land, llama.cpp closed the 2.6 request as not_planned
Jul 31
- Review testing on Ling 3.0 flash - From one prompt to a 3D world
Aug 4
- inclusionAI/Ling-3.0-flashhuggingface.co
Aug 5
- Ling-3.0-flash MXFP4 released and running locally on one DGX Spark.
Aug 8
- Ant Group has just released Ling 3.0 Flash, a 124B open weights model that scores 38 on the...
Aug 9
- Two flags took the official Ling-3.0-flash INT4 from 20.8 to 38.7 tok/s on one DGX Spark
Aug 11
- Ling-3.0-flash quant ladder on one DGX Spark: the whole thing sits in a 32 to 40 tok/s band
Aug 22
- AntLing released a dspark draft model for Ling-3.0-flash

Daily brief

Get tomorrow's AI brief in your inbox

Ant Group releases Ling 3.0 Flash, a 124B open-weights MoE model

How this story unfolded

Daily brief

More stories today