Model Anatomy See what's inside open-weight AI models

Can I run Nanbeige4.3-33B on a Radeon RX 7900 XTX?

Yes, with a GGUF version

The published weights are too big, but the IQ4_XS file from AtomicChat/Nanbeige4.3-33B-GGUF (20.3 GB) fits, with room for ≈ 6K tokens of context.

Full analysis of Nanbeige4.3-33B Everything that runs on Radeon RX 7900 XTX Get the IQ4_XS file on Hugging Face
ollama run hf.co/AtomicChat/Nanbeige4.3-33B-GGUF:IQ4_XS

Files and context on Radeon RX 7900 XTX

Each file of Nanbeige4.3-33B (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

FileSizeOn Radeon RX 7900 XTX Context (16-bit)Context (8-bit)
As published66.2 GBDoesn't fit——
BF16bartowski/Nanbeige_Nanbeige4.3-33B-GGUF66.2 GBDoesn't fit——
Q8_0bartowski/Nanbeige_Nanbeige4.3-33B-GGUF35.2 GBDoesn't fit——
Q6_K_Lbartowski/Nanbeige_Nanbeige4.3-33B-GGUF30.4 GBDoesn't fit——
Q6_Kbartowski/Nanbeige_Nanbeige4.3-33B-GGUF29.0 GBDoesn't fit——
Q6_K_Sbartowski/Nanbeige_Nanbeige4.3-33B-GGUF27.8 GBDoesn't fit——
Q5_K_Mbartowski/Nanbeige_Nanbeige4.3-33B-GGUF25.1 GBDoesn't fit——
Q5_K_Sbartowski/Nanbeige_Nanbeige4.3-33B-GGUF23.7 GBDoesn't fit——
Q4_K_Lbartowski/Nanbeige_Nanbeige4.3-33B-GGUF22.3 GBDoesn't fit——
Q4_1bartowski/Nanbeige_Nanbeige4.3-33B-GGUF21.2 GBDoesn't fit2K4K
IQ4_NLbartowski/Nanbeige_Nanbeige4.3-33B-GGUF20.9 GBDoesn't fit3K7K
Q4_K_Mbartowski/Nanbeige_Nanbeige4.3-33B-GGUF20.8 GBDoesn't fit4K7K
Q4_K_Sbartowski/Nanbeige_Nanbeige4.3-33B-GGUF19.7 GBTight9K17K
Q4_0bartowski/Nanbeige_Nanbeige4.3-33B-GGUF19.3 GBTight10K21K
IQ4_XSbartowski/Nanbeige_Nanbeige4.3-33B-GGUF18.6 GBTight14K27K
IQ3_Mbartowski/Nanbeige_Nanbeige4.3-33B-GGUF17.8 GBTight17K35K
Q3_K_Lbartowski/Nanbeige_Nanbeige4.3-33B-GGUF17.3 GBFits20K40K
Q3_K_Mbartowski/Nanbeige_Nanbeige4.3-33B-GGUF15.9 GBFits26K52K
IQ3_XSbartowski/Nanbeige_Nanbeige4.3-33B-GGUF15.2 GBFits29K59K
Q3_K_Sbartowski/Nanbeige_Nanbeige4.3-33B-GGUF15.2 GBFits29K59K
IQ3_XXSbartowski/Nanbeige_Nanbeige4.3-33B-GGUF14.5 GBFits33K65K
Q2_Kbartowski/Nanbeige_Nanbeige4.3-33B-GGUF13.5 GBFits37K74K
IQ2_Mbartowski/Nanbeige_Nanbeige4.3-33B-GGUF12.2 GBFits43K86K
IQ2_Sbartowski/Nanbeige_Nanbeige4.3-33B-GGUF11.2 GBFits48K96K
IQ2_XSbartowski/Nanbeige_Nanbeige4.3-33B-GGUF10.6 GBFits50K101K
IQ2_XXSbartowski/Nanbeige_Nanbeige4.3-33B-GGUF10.1 GBFits53K106K
Q8_0AtomicChat/Nanbeige4.3-33B-GGUF35.2 GBDoesn't fit——
Q6_KAtomicChat/Nanbeige4.3-33B-GGUF30.1 GBDoesn't fit——
Q5_KAtomicChat/Nanbeige4.3-33B-GGUF27.9 GBDoesn't fit——
Q5_K_MAtomicChat/Nanbeige4.3-33B-GGUF24.7 GBDoesn't fit——
Nanbeige4.3-33B-Q5_K_MAtomicChat/Nanbeige4.3-33B-GGUF23.5 GBDoesn't fit——
Nanbeige4.3-33B-AD-Q5_K_M-Q4_K_MAtomicChat/Nanbeige4.3-33B-GGUF22.6 GBDoesn't fit——
Q4_K_MAtomicChat/Nanbeige4.3-33B-GGUF21.2 GBDoesn't fit2K4K
IQ4_XSAtomicChat/Nanbeige4.3-33B-GGUF20.3 GBTight6K11K
IQ3_SAtomicChat/Nanbeige4.3-33B-GGUF17.1 GBFits21K42K
IQ3_XXSAtomicChat/Nanbeige4.3-33B-GGUF14.9 GBFits31K61K
IQ2_SAtomicChat/Nanbeige4.3-33B-GGUF13.7 GBFits36K72K

Nanbeige4.3-33B on other hardware

Popular models that run on Radeon RX 7900 XTX

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.