Unverified90% confidenceFactTime unknown
千问（Qwen）和DeepSeek都提供了基于MOE架构的版本
1
Sources
90%
Confidence
Long-term
Relevance
6/2/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOpen-source large model families such as Qwen, Mixtral, and DeepSeek all adopt MoE designs.81% similarUnverifiedDeepSeek officially provides multiple distilled versions based on the Qwen and Llama architectures, with parameter counts ranging from 1.5B to 70B.79% similarUnverifiedThe Qwen2.5 series (including Math variants) and DeepSeek both provide comprehensive tool-calling support and support local deployment.79% similarUnverifiedAlibaba Lingma supports Qwen and DeepSeek series models and entered the market relatively late.75% similarVerifiedUnsloth supports fine-tuning mainstream open-source large models including Gemma 4, Qwen3, DeepSeek, and gpt-oss.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/38094API
curl https://kongchang.com/api/v1/knowledge/claims/38094MCP
get_claim(id=38094)