Expired50% confidenceReleaseExact time
Weibo AI released the open-source model VibeThinker 3B, with the company officially claiming this 3B model reaches flagship large model performance on multiple benchmarks, though the community has raised concerns about overfitting.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/8/2026
First Seen
Valid until: 10/6/2026(expired)
Sources
Related Claims
Unverified腾讯正式发布了开源权重大模型 HY3，是预览版之后的完整推理模型66% similarUnverified阿里巴巴在周日发布了升级版通义千问（Qwen）模型预览版，跟进 Kimi K3 的发布62% similarUnverifiedOpenAI, Anthropic, and Google have each formed a three-tier model lineup comprising a flagship reasoning model, a balanced model, and a lightweight model.56% similarUnverifiedGoogle released the Gemma 4 series of open-source models, comprising four models of different parameter scales.53% similarUnverifiedMoE概念最早由Jacobs等人于1991年提出，谷歌2017年将其规模化至语言模型，Mistral的Mixtral系列将稀疏MoE带入开源社区53% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/212926API
curl https://kongchang.com/api/v1/knowledge/claims/212926MCP
get_claim(id=212926)