Ant Ling Releases Ling-3.1-flash Model: ~560B Total Parameters, 25B Activated Parameters
Summary
Ant Group's Ling team released the Ling-3.1-flash model, with roughly 560B total parameters and about 25B activated per token, and a context window of up to 1M. The company says it is optimized for general agents, search, everyday office work and software development, and will move from a two-week free trial to a paid service, at which point it plans to open the full 1M context and release the model as open source.
Bubble analysis
This is a model release — a technical milestone rather than a flow of money — so it says little directly about whether AI investment is overheating. Its only link to the bubble question is that capabilities keep improving fast and the model is being given away free to win users, hinting vendors are still burning cash for market share, but the article carries no spending or revenue figures to judge that.
Related signals
linked by shared tags
- Huawei Open-Sources Pretraining, SFT and Post-Training RL Code for openPangu-2.0#open-source#llm
- StepFun Releases Flagship Model Step 5 Preview: Enters Global Top Three Open-Source Models on AA Leaderboard, Open-Sourcing on October 15#open-source#llm
- StepFun Releases New Flagship Model Step 5 Preview#open-source#llm
- China Telecom Open-Sources Xing4.0-29B-A4B, Its First Fully Domestic-Stack Lightweight Agent LLM#open-source#llm
- Bilibili Open-Sources the Index-Translate Multilingual Translation Model Family, Text Models Cover 150 Languages#open-source