Edge0-145B-NVFP4

Try Try

Edge0-145B-NVFP4 has 4.95B parameters and takes 5.98 GB on disk. See where its parameters live, layer by layer, and what hardware it needs.

At a glance

Edge0/Edge0-145B-NVFP4
Repo createdNov 11, 2026updated Nov 15, 2026
Model type—Not shown: breakdown incomplete
InputsText + imagesvision encoder in a separate mmproj file, not counted here
Total parameters4.95B
Active per token—Not shown: breakdown incomplete
Experts—Not shown: breakdown incomplete
Max context (from config)256K tokens
Layers4836 linear + 12 full attention
On disk5.98 GB1 file
PrecisionQ8_0 (74%), BF16 (26%), other (0.17%)
Quantizationmodel-skeleton (GGUF)Q8_0 · blocks of 32 · 85% of parameters · unclassified kept in BF16
Fewest GPUs1× RTX 4060a single 8 GB card · weights only
FamilyQuantization of Qwen/Qwen3.8-Flash-Next
Licenseedge0-community-model-1.0 (custom)
GitHubEdge0-AI/Edge0