Unverified50% confidenceFactExact time
The Ling-3.0-flash model has a total parameter count of 124B, with approximately 5.1B parameters activated per token, employs a KDA + MLA hybrid attention mechanism, and has a context length of 262k.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/5/2026
First Seen
Valid until: 11/3/2026
Sources
中国四大AI实验室路线对比：Qwen、DeepSeek、Moonshot、Ling战略解析
Related Claims
UnverifiedLing-3.0-flash adopts a Mixture of Experts (MoE) architecture with an activation ratio of only approximately 4% (approximately 5.1B activated out of 124B total parameters).78% similarUnverified混元3是混合专家模型（MoE），总参数量为295B，激活参数约21B65% similarUnverified千问3.6的35B MOE模型每次只激活其中3B的参数64% similarUnverifiedXiaomi MiMo adopts a Mixture-of-Experts (MoE) architecture with a total parameter count of 24.6B, activating only approximately 5.3B parameters per inference.64% similarUnverifiedHy3 adopts a Mixture of Experts (MoE) architecture with a total parameter count of 295 billion (295B), activating only 21 billion (21B) parameters per inference.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/687534API
curl https://kongchang.com/api/v1/knowledge/claims/687534MCP
get_claim(id=687534)