Can T4 16GB run Qwen3.6-35B-A3B?

NVIDIA · 16 GB GDDR6 · 320 GB/s bandwidth · Alibaba · 36.0B (active 3.0B) · MoE · model ctx up to 256K

No — Qwen3.6-35B-A3B at Q4 needs ≈ 21.1 GB even at 4K context, beyond the T4 16GB's 16 GB GDDR6. It would only run with heavy CPU/disk offload. Decode at Q4/8K: ≈ 108.4 tok/s (estimated, batch 1). FP16 needs ≈ 73.6 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.

Fit & speed by quant and context

Quant	Context	VRAM needed	Fits?	tok/s (decode)
Q4	4K	21.1 GB	No	111.5 est.
Q4	8K default	21.4 GB	No	108.4 est.
Q4	32K	23.2 GB	No	92.2 est.
FP16	4K	73.3 GB	No	—
FP16	8K default	73.6 GB	No	—
FP16	32K	75.4 GB	No	—

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	T4 16GB VRAM
Q4	19.8 GB	0.6 GB	1 GB	21.4 GB	16 GB
FP16	72.0 GB	0.6 GB	1 GB	73.6 GB	16 GB

Other GPUs that run Qwen3.6-35B-A3B

• B200 192GB SXM — ≈1697.0 tok/s Q4 est.
• MI300X 192GB — ≈1124.2 tok/s Q4 est.
• H200 141GB SXM5 — ≈1018.2 tok/s Q4 est.
• H100 80GB SXM5 — ≈710.6 tok/s Q4 est.
• Gaudi 3 128GB — ≈694.9 tok/s Q4 est.
• A100 80GB SXM4 — ≈432.5 tok/s Q4 est.
• RTX 5090 32GB — ≈380.1 tok/s Q4 est.
• RTX 4090 24GB — ≈213.8 tok/s Q4 est.

All GPUs for Qwen3.6-35B-A3B →

Other models for the T4 16GB

• MiniCPM5-1B — ≈188.6 tok/s Q4 est.
• Qwen3.5-2B — ≈101.8 tok/s Q4 est.
• Llama-3.2-3B-Instruct — ≈63.6 tok/s Q4 est.
• gpt-oss-20b — ≈56.6 tok/s Q4 est.
• Qwen3-4B-Instruct-2507 — ≈50.9 tok/s Q4 est.
• Qwen3-4B-Thinking-2507 — ≈50.9 tok/s Q4 est.
• NVIDIA-Nemotron-3-Nano-4B-BF16 — ≈50.9 tok/s Q4 est.
• gemma-4-E2B-it — ≈39.9 tok/s Q4 est.

All models on the T4 16GB →