Unverified50% confidenceOpinionExact time
MiMo-V2.6的具体基准测试成绩及与同类开源全模态模型（如Qwen-VL系列）的横向对比仍有待检验
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/22/2026
First Seen
Valid until: 12/21/2026
Sources
Related Entities
Related Claims
UnverifiedMeta的Llama 3.1支持405B参数，与Qwen2.5、DeepSeek-V3等开源模型在多项基准测试中已接近甚至超越部分闭源商业模型68% similarUnverifiedB站开发者Luke对Unsloth发布的Qwen3系列27B模型进行了从Q1到Q8多个量化级别的横评测试67% similarUnverifiedLFM2.5在多项基准测试中展现出与体量高达其4倍模型相当的竞争力67% similarUnverified该测试为单一来源的个人测试，模型型号被标注为Qwen 3.8 27B存在命名歧义，实际配置、量化方式、生成耗时等细节均未披露，可复现性有待验证66% similarUnverified测试基于Unsloth发布的Qwen3 27B NVFP4量化版本进行66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/940613API
curl https://kongchang.com/api/v1/knowledge/claims/940613MCP
get_claim(id=940613)