Unverified50% confidenceFactExact time
AdaptiveSpec支持DeepSeek-R1-Distill-Llama-8B、Llama-3.1-8B-Instruct和Qwen3-8B三个主流目标模型
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/7/2026
First Seen
Valid until: 12/6/2026
Sources
Related Entities
Related Claims
UnverifiedQwen 3.8 Flash Next在某些任务上可与DeepSeek R1 Pro相媲美67% similarUnverified参与本轮横评的模型包括Fable 5、DXU E4GA、DXU E4 Slash N731正式版、千问三代MiniMax 0720和DeepSeek V4 Pro67% similarUnverified寒武纪已完成GLM、DeepSeek、Qwen、Kimi、MiniMax五款国产主流大模型的推理适配，第六代芯片架构正在研发中66% similarUnverifiedUnsloth Desktop当前已支持Qwen3、Meta系列模型以及DeepSeek等多种主流模型65% similarUnverifiedQwythos-9B建立在未经审查的Qwen 3.5基础模型之上65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/869121API
curl https://kongchang.com/api/v1/knowledge/claims/869121MCP
get_claim(id=869121)