Updated

Ling-3.1-flash

InclusionAI released Ling-3.1-flash, a hybrid reasoning model with 560B total parameters, about 25B active per token, and a model context limit of 1M tokens. It builds on Ling-3.0-flash and targets real-world agent, search, office, and software-engineering tasks.

By No Code MBA · Facts checked

Provider
InclusionAI
Announcement date
Status at last check
Generally available

What changed

InclusionAI describes Ling-3.1-flash as an upgrade to Ling-3.0-flash, with a hybrid-linear MoE architecture, 560B total and 25B active parameters per token, and a 1M-token model context limit. The model card reports improvements across general-purpose agents, search, office productivity, software engineering, finance, and materials-science tasks.

Access and availability

Download the model from the official InclusionAI Hugging Face repository. Ant's official Ling service documentation states that the hosted service currently offers up to 256K context, while the model itself supports up to 1M.

Official access documentation

What you can build

For local deployment, start with the model card's hardware and serving guidance, then benchmark representative long-context tasks before choosing a hosted or self-managed setup.

Limitations

The hosted Ling service currently offers up to 256K context even though the model card specifies a 1M model limit. The model card's evaluations and performance claims are from the provider; local deployment requires compatible serving software and substantial accelerator resources.

Sources and release history