Today for AI
HOT RADAR
llmHEAT 11.7°

AI CURATED HOT STORY · 10/15/2026

Ant Group Open-Sources Ling-3.1-flash: A 560B MoE Model with 1M Context for Real-World Workflows

1 reports archived1 independent sourcesupdated 10/15/2026, 19:47:57
Synthesis & Latest Updates
1 Sources Cross-Validated

Ant Group's Bailing team has open-sourced Ling-3.1-flash, a hybrid linear architecture model featuring 560B total parameters, 25B active parameters, and support for a 1M context window. The model demonstrates strong performance in benchmarks like GDPVal-AA, HealthBench, and FrontierSWE, targeting real-world business scenarios such as office automation, medical assistance, and complex software engineering tasks.

LATEST/Ant Group's Bailing team has open-sourced Ling-3.1-flash, a hybrid linear architecture model featuring 560B total parameters, 25B active parameters, and support for a 1M context window. The model demonstrates strong performance in benchmarks like GDPVal-AA, HealthBench, and FrontierSWE, targeting real-world business scenarios such as office automation, medical assistance, and complex software engineering tasks.

HEAT TRENDHourly heat curve

3 fully observed hours
Now
11.7
Peak
11.920:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

TIMELINECoverage timeline

Total 1 reports · Latest first
  1. InfoQ 中文T2·78 pts
    • Ling-3.1-flash utilizes a hybrid linear architecture to achieve efficient inference with only 25B active parameters out of 560B total.
    • The model natively supports a 1M context window, enhancing its ability to handle long documents, codebases, and complex multi-turn interactions.
    • It achieves high scores across office, medical, and software engineering benchmarks, emphasizing 'real-world usability' over leaderboard chasing.