Can L40S 48GB run Step-3.5-Flash?

NVIDIA · 48 GB GDDR6 · 864 GB/s bandwidth · StepFun · 196.8B (active 11.0B) · MoE · model ctx up to 256K

No — Step-3.5-Flash at Q4 needs ≈ 110.34 GB even at 4K context, beyond the L40S 48GB's 48 GB GDDR6. It would only run with heavy CPU/disk offload. Decode at Q4/8K: ≈ 26.5 tok/s (estimated, batch 1). FP16 needs ≈ 396.8 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	110.34 GB	No	27.0 est.
Q4	8K default	111.44 GB	No	26.5 est.
Q4	32K	118.04 GB	No	23.6 est.
FP16	4K	395.7 GB	No	—
FP16	8K default	396.8 GB	No	—
FP16	32K	403.4 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	L40S 48GB VRAM
Q4	108.24 GB	2.2 GB	1 GB	111.44 GB	48 GB
FP16	393.6 GB	2.2 GB	1 GB	396.8 GB	48 GB

Other GPUs that run Step-3.5-Flash

• B200 192GB SXM — ≈462.8 tok/s Q4 est.
• MI300X 192GB — ≈306.6 tok/s Q4 est.
• H200 141GB SXM5 — ≈277.7 tok/s Q4 est.
• Gaudi 3 128GB — ≈189.5 tok/s Q4 est.
• Mac Studio M2 Ultra 192GB — ≈46.3 tok/s Q4 est.
• MacBook Pro M4 Max 128GB — ≈31.6 tok/s Q4 est.
• MacBook Pro M3 Max 128GB — ≈23.1 tok/s Q4 est.

All GPUs for Step-3.5-Flash →

Other models for the L40S 48GB

• MiniCPM5-1B — ≈509.1 tok/s Q4 est.
• Qwen3.5-2B — ≈274.9 tok/s Q4 est.
• GLM-4.7-Flash — ≈183.3 tok/s Q4 est.
• Qwen3-30B-A3B-Instruct-2507 — ≈183.3 tok/s Q4 est.
• Qwen3-Coder-Next — ≈183.3 tok/s Q4 est.
• Qwen3.5-35B-A3B — ≈183.3 tok/s Q4 est.
• Qwen3.6-35B-A3B — ≈183.3 tok/s Q4 est.
• Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 — ≈183.3 tok/s Q4 est.

All models on the L40S 48GB →