Expired50% confidenceReleaseExact time
The developer provides five quantization formats: INT8 ConvRot (4.5 GB), FP8 E4M3 (4.2 GB), bf16 (8.3 GB), NVFP4 (2.9 GB), and MXFP8 (4.7 GB).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/20/2026
First Seen
Valid until: 10/18/2026(expired)
Sources
Related Claims
UnverifiedCalculated from a storage perspective: 27B parameters × 2 bytes (FP16) ≈ 54 GB, INT8 ≈ 27 GB, INT4 ≈ 13.5 GB, 2-bit ≈ 6.75 GB, 1.5-bit approaches 5 GB; fitting within 4 GB means the average bit width needs to be below 1.5-bit.80% similarUnverifiedUnder hardware conditions of 16 GB of memory, only models within 5 GB can be accommodated, so the developer used the Qwen 2.5 7B 4-bit quantized model, which is approximately 4.8 GB in size, with an 8K context window.78% similarVerified8 GB of VRAM is recommended for running 7B models (Q4 quantization), 16 GB for 7B to 14B models, 24 GB for 14B to 32B models, and 32 GB for 32B to 70B models (Q4 quantization)77% similarUnverifiedThe Qwen 3.5 9B model is approximately 6.6 GB and is suited for GPUs with 8 GB of VRAM75% similarUnverifiedM3原始权重存储下限：BF16约854GB，8-bit约428GB，4-bit约214GB74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/568120API
curl https://kongchang.com/api/v1/knowledge/claims/568120MCP
get_claim(id=568120)