Unverified50% confidenceReleaseExact time
蚂蚁集团实验室发布了 Ling 3.0 Tiny 模型，总参数7.9B，每token激活约1.3B
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/10/2026
First Seen
Valid until: 11/8/2026
Sources
Related Claims
UnverifiedLing 3.0 Tiny 每处理一个token的算术运算量约为8B稠密模型的六分之一72% similarUnverifiedLing-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).62% similarUnverifiedMicrosoft's Phi series (Phi-1 to Phi-3), with parameter counts ranging from 1.3B to 3.8B, are representative works in the area of small language models.58% similarUnverifiedPhi-3-mini（3.8B参数）在部分基准测试上超越了Mixtral 8x7B等更大的混合专家模型57% similarUnverifiedTongyi Qianwen 3.7 Max is a large language model launched by Alibaba Cloud and is a high-performance version of the Qianwen series.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/723986API
curl https://kongchang.com/api/v1/knowledge/claims/723986MCP
get_claim(id=723986)