Can RTX 5080 16GB run Qwen3.5-122B-A10B-Heretic-v2-MLX-mixed-3.8bit?

NVIDIA · 16 GB GDDR7 · 960 GB/s bandwidth · TheCluster · 19.8B (active 19.8B) · MoE · model ctx up to NoneK

CUDA

Yes — the RTX 5080 16GB (16 GB GDDR7) can run Qwen3.5-122B-A10B-Heretic-v2-MLX-mixed-3.8bit. At Q4 with the default 8K context it needs ≈ 15.85 GB VRAM (weights 10.89 GB + KV cache 3.96 GB + 1 GB runtime overhead), which fits within 16 GB. Decode at Q4/8K: ≈ 30.9 tok/s (estimated, batch 1). FP16 needs ≈ 44.56 GB — does not fit on this card. Estimates come from memory-bandwidth math; rows tagged measured override estimates. Relative ranking is more reliable than absolute tok/s. Methodology. Buy RTX 5080 16GB on Amazon →

Fit & speed by quant and context

Quant	Context	VRAM needed	Fits?	tok/s (decode)
Q4	4K	13.87 GB	Yes	30.9 ± est.
Q4	8K default	15.85 GB	Yes	30.9 ± est.
Q4	32K	27.73 GB	No	29.3 ± est.
FP16	4K	42.58 GB	No	3.4 ± est.
FP16	8K default	44.56 GB	No	3.1 ± est.
FP16	32K	56.44 GB	No	1.9 ± est.

Fits = weights + KV(ctx) + 1 GB overhead ≤ 95% of VRAM. Measured anchors are context-agnostic; the fit verdict is recomputed per context. A missing tok/s means the model is far beyond this card (offload-only territory).

VRAM breakdown at 8K context

Quant	Weights	KV cache	Overhead	Total needed	RTX 5080 16GB VRAM
Q4	10.89 GB	3.96 GB	1 GB	15.85 GB	16 GB
FP16	39.6 GB	3.96 GB	1 GB	44.56 GB	16 GB

Other GPUs that run Qwen3.5-122B-A10B-Heretic-v2-MLX-mixed-3.8bit

• B200 192GB SXM — ≈257.1 tok/s Q4 est.
• MI300X 192GB — ≈170.3 tok/s Q4 est.
• H200 141GB SXM5 — ≈154.3 tok/s Q4 est.
• H100 80GB SXM5 — ≈107.7 tok/s Q4 est.
• Gaudi 3 128GB — ≈105.3 tok/s Q4 est.
• A100 80GB SXM4 — ≈65.5 tok/s Q4 est.
• RTX 5090 32GB — ≈57.6 tok/s Q4 est.
• RTX 4090 24GB — ≈32.4 tok/s Q4 est.

All GPUs for Qwen3.5-122B-A10B-Heretic-v2-MLX-mixed-3.8bit →

Other models for the RTX 5080 16GB

• MiniCPM5-1B — ≈565.7 tok/s Q4 est.
• Qwen3.5-2B — ≈305.5 tok/s Q4 est.
• DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx — ≈265.6 tok/s Q4 est.
• gpt-oss-20b — ≈204.9 tok/s Q4 measured
• Llama-3.2-3B-Instruct — ≈190.9 tok/s Q4 est.
• Mistral-7B-Instruct-v0.3 — ≈163.6 tok/s Q4 measured
• gemma-4-26B-A4B-it — ≈160.8 tok/s Q4 est.
• Qwen3-4B-Instruct-2507 — ≈152.7 tok/s Q4 est.

All models on the RTX 5080 16GB →