Verified70% confidenceFactTime unknown
Open-source multimodal models LLaVA, InternVL, and Qwen-VL typically adopt a 'vision encoder + projection layer + language model' architecture.
4
Sources
70%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
GitHub 8000+ Star：awesome-LLM-resources最全大语言模型资源库解析
githubWangRongsheng
Related Entities
Related Claims
Unverified开源多模态模型包括LLaVA、MiniCPM-V、InternVL、Qwen-VL等80% similarUnverifiedMainstream open-source models focused on document understanding, such as InternVL2 and Qwen-VL, mostly have parameter counts in the range of 3B to 72B.72% similarUnverifiedLocal large models typically refer to open-source LLMs (such as Llama and Qwen series) hosted by tools like Ollama and LM Studio, responsible for text-side tasks such as script generation, subtitle writing, and intelligent scheduling71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/31733API
curl https://kongchang.com/api/v1/knowledge/claims/31733MCP
get_claim(id=31733)