Unverified85% confidenceFactTime unknown
Gemma 4引入了多模态能力，Qwen3支持超长上下文，DeepSeek采用MoE（混合专家）架构
1
Sources
85%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Unsloth：本地微调大模型的高效开源工具
githubunslothai
Related Entities
Related Claims
UnverifiedGemma 4、Qwen 2.5、Kimi K2.5等新发布的模型都是视觉语言模型（VLM），越来越多模型在发布首日就具备视觉能力70% similarUnverifiedv1.0.0 将张量并行支持扩展至大多数主流模型，其中包括 Gemma466% similarUnverifiedQwen3.7 Max对LangChain、CrewAI、Hermes等主流智能体框架具有广泛兼容性65% similarUnverified据社区讨论，DeepSeek V4将带来原生MXFP4混合专家（MoE）架构，配合超大上下文能力65% similarUnverifiedQwen2.5、Gemma3 等系列均已推出专为边缘场景优化的 1B-3B 量化版本65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/32191API
curl https://kongchang.com/api/v1/knowledge/claims/32191MCP
get_claim(id=32191)