DeepSeek-V4-Flash-180B

Deepseek family · 158.0B params

1recorded run

80.8best tok/s

1verified

Evidence: 1 sourced run · 1 independent source · first recorded Jul 18, 2026. What these labels mean.

What the DeepSeek-V4-Flash-180B numbers mean

The fastest verified result on our board is 80.8 output tok/s on an Intel Arc Pro B70 x4 at FP8 with a 256-token context. That is about 12.4 ms per generated token — 500 tokens takes roughly 6.2 seconds.

How much VRAM does DeepSeek-V4-Flash-180B need?

DeepSeek-V4-Flash-180B has 158.0B parameters. These estimates use the same formula as the VRAM calculator: weights, plus about 15% for a moderate-context KV cache and runtime overhead, plus a small engine floor.

Quantisation	Estimated VRAM	Smallest card that fits
Q4_K_M (4-bit)	103.0 GB	141 GB
Q8_0 (8-bit)	193.9 GB	multi-GPU or CPU offload
FP16 (16-bit)	364.2 GB	multi-GPU or CPU offload

Long contexts and large batches raise the real number — read the smallest card as "possible with a reduced context". Measured throughput is on the benchmark board.

Fastest GPUs for this model (best tok/s)

Intel Arc Pro B70 x480.8

Models with a similar footprint

Model	Params	Q4_K_M weights
CYBER-FROST-3.8	180B	101.2 GB
Qwen3.8-Flash-Next	180B	101.2 GB
Qwen3.8-Flash-Next-hibrid48	180B	101.2 GB
Qwen3.8-Flash-Next-REAP-256-duo	180B	101.2 GB
Ling-3.0-flash	127B	71.4 GB
Qwen 3.8 Flash	125B	70.3 GB

Ranked by distance from DeepSeek-V4-Flash-180B's parameter count. Full benchmark board.

All benchmarks Methodology

GPU	Model	Quant	Backend	Ctx	tok/s	prompt tok/s	VRAM GB	Source
Intel Arc Pro B70 x4 · 32 GB	DeepSeek-V4-Flash-180B	FP8	vllm 0.1.dev1172+g4a6fd8747.xpu; local vL…	256	80.8 ✓	—	—	src