Unverified50% confidenceFactExact time
All of these Qwen3 uncensored models provide GGUF format support and retain native MTP (Multi-Token Prediction) capability.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/7/2026
First Seen
Valid until: 12/6/2026
Sources
Related Entities
Related Claims
UnverifiedThe Qwen3.8-27B model is available in multiple formats including Safetensors, GGUF, NVFP4, NVFP4 GGUF variants, and GPTQ-Int4 quantized versions.78% similarUnverifiedThe Qwen3 series natively supports MTP.78% similarUnverifiedQwythos-9B is built on top of the uncensored Qwen 3.5 base model.70% similarUnverifiedOllama natively supports the GGUF format, and most models have been pre-quantized to specifications such as Q4_K_M or Q8_0.69% similarUnverifiedGQA（分组查询注意力）等技术可以有效压缩KV Cache，Qwen3系列也采用了类似优化69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/871909API
curl https://kongchang.com/api/v1/knowledge/claims/871909MCP
get_claim(id=871909)