RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can Qwen3.5 4B run on the RX 7900 GRE?

Yes. At 8K context the best fit is F16, using about 9.0 GB of the 15.2 GB usable on the RX 7900 GRE. Estimated speed ceiling: ~68 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K)2.6 GB Fits2.8 GB Fits3.5 GB Fits6.5 GB Fits330
3-bit (Q3_K_M) •3.1 GB Fits3.3 GB Fits4.0 GB Fits7.0 GB Fits251
4-bit (Q4_K_M) •3.6 GB Fits3.7 GB Fits4.4 GB Fits7.4 GB Fits210
5-bit (Q5_K_M) •3.9 GB Fits4.1 GB Fits4.8 GB Fits7.8 GB Fits183
6-bit (Q6_K) •4.3 GB Fits4.4 GB Fits5.2 GB Fits8.2 GB Fits163
8-bit (Q8_0) •5.2 GB Fits5.3 GB Fits6.0 GB Fits9.0 GB Fits129
16-bit (F16/BF16) •8.8 GB Fits9.0 GB Fits9.7 GB Fits12.7 GB Fits68

• real GGUF file size from unsloth/Qwen3.5-4B-GGUF. Speed is a bandwidth ceiling (576 GB/s), not a benchmark. Methodology.

More

Qwen3.5 4B on every GPU and Mac · Every model that runs on the RX 7900 GRE · Other models that fit 16 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (Qwen/Qwen3.5-4B), AMD specs. Data as of 2026-11-17.