inclusionAI/Ling-3.0-flash · Hugging Face

The Ling-3.0-flash MoE is now open-weighted at 124B A5B params. I know the original announcements were before the Kimi K3, DeepSeek-V4-Flash and Qwen3.8 hype, but this model might still have a good niche for itself due to its sizing. Discussion on the benchmarks are here: https://www.reddit.com/r/L

中文全文翻译正在补齐，可先查看下方英文原文。

The Ling-3.0-flash MoE is now open-weighted at 124B A5B params. I know the original announcements were before the Kimi K3, DeepSeek-V4-Flash and Qwen3.8 hype, but this model might still have a good niche for itself due to its sizing. Discussion on the benchmarks are here: https://www.reddit.com/r/L