Speed test matrix

Qwen3.5-4B-AWQ-4bit on RTX 3060 12GB

vllm · 0.32.0 · AWQ 4-bit6 contexts (0–176K) × 8 concurrency levels (1–128) · prefill at 8K, 16K, 32K, 64K, 128K

by @Sandbergllm-inference-bench v0.8.0Measured Oct 30, 2026, 21:30 UTC

Hardware

Type: Discrete GPU
GPU: RTX 3060
VRAM: 12 GB
CPU: Intel Xeon E5-2680 v4 (14C)
RAM: 32 GB
OS: Ubuntu 24.04.5

Engine

Engine: vllm
Version: 0.32.0
Quantization: AWQ 4-bit
Backend: cuda
Repository: github.com/vllm-project/vllm
Build: vllm/vllm-openai:v0.32.0-x86_64

Run

Served model: bench
Revision: main
Benchmark tool: llm-inference-bench v0.8.0
Decode test: Fixed duration
Prefill mode: Measured during decode warm-up
Time per measurement: 30 s
Max output: 8,192 tokens

Launch command

# Untuned out-of-the-box baseline. Spec decode: none
# 1x RTX 3060 12GB, stock 170 W limit, PCIe Gen3 x16 | Intel Xeon E5-2680 v4 14C | 32 GB DDR4-2400 | Ubuntu 24.04.5, NVIDIA driver 580.178.04
vllm serve cyankiwi/Qwen3.5-4B-AWQ-4bit --max-model-len 188416 --gpu-memory-utilization 0.94 --max-num-seqs 128 --enable-prefix-caching --trust-remote-code --language-model-only

Prefill

prompt tok/s by context length

8K context2,7652,947 ms TTFT

16K context2,5736,264 ms TTFT

32K context2,29313,998 ms TTFT

64K context1,85434,540 ms TTFT

128K context1,34395,295 ms TTFT

Sustained decode

total tok/s · rows are context length, columns are simultaneous requests

Context	1	2	4	8	16	32	64	128
0	83.6107 ms TTFT	160.196 ms TTFT	295.8210 ms TTFT	515.6346 ms TTFT	812.4583 ms TTFT	Over capacity	Over capacity	Over capacity
16K	73.0265 ms TTFT	125.9359 ms TTFT	197.8535 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
32K	64.7423 ms TTFT	103.2571 ms TTFT	147.5889 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
64K	52.7471 ms TTFT	75.7704 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
128K	38.4844 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity
176K	31.81,008 ms TTFT	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity	Over capacity

Over capacity: not run because this many requests at this context would exceed the server's KV cache.

TTFT is measured after the benchmark warms the server's prompt cache, so it mostly excludes prefill.

All measurements

Latencies are medians. Total tok/s is output across all simultaneous requests.

Phase	Context	Requests at once	Status	Total tok/s	Per request tok/s	TTFT	Time between tokens	Average active requests	Completed requests	Output tokens	Measured for
Prefill	8K	1	Valid	2,765.0	—	2,947 ms	—	—	1	—	—
Prefill	16K	1	Valid	2,573.0	—	6,264 ms	—	—	1	—	—
Prefill	32K	1	Valid	2,293.0	—	13,998 ms	—	—	1	—	—
Prefill	64K	1	Valid	1,854.0	—	34,540 ms	—	—	1	—	—
Prefill	128K	1	Valid	1,343.0	—	95,295 ms	—	—	1	—	—
Sustained decode	0	1	Valid	83.6	83.9	107 ms	11.9 ms	1.0	0	2,509	30.0 s
Sustained decode	0	2	Valid	160.1	80.4	96 ms	12.4 ms	2.0	0	4,804	30.0 s
Sustained decode	0	4	Valid	295.8	74.5	210 ms	13.4 ms	4.0	0	8,872	30.0 s
Sustained decode	0	8	Valid	515.6	65.2	346 ms	15.3 ms	8.0	0	15,464	30.0 s
Sustained decode	0	16	Valid	812.4	51.6	583 ms	19.4 ms	16.0	0	24,368	30.0 s
Sustained decode	0	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	0	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	1	Valid	73.0	73.2	265 ms	13.7 ms	1.0	0	2,190	30.0 s
Sustained decode	16K	2	Valid	125.9	63.0	359 ms	15.9 ms	2.0	0	3,776	30.0 s
Sustained decode	16K	4	Valid	197.8	49.3	535 ms	20.3 ms	4.0	0	5,933	30.0 s
Sustained decode	16K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	16K	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	1	Valid	64.7	64.9	423 ms	15.4 ms	1.0	0	1,942	30.0 s
Sustained decode	32K	2	Valid	103.2	51.6	571 ms	19.4 ms	2.0	0	3,096	30.0 s
Sustained decode	32K	4	Valid	147.5	36.6	889 ms	27.3 ms	4.0	0	4,424	30.0 s
Sustained decode	32K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	32K	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	64K	1	Valid	52.7	52.9	471 ms	18.9 ms	1.0	0	1,582	30.0 s
Sustained decode	64K	2	Valid	75.7	37.9	704 ms	26.4 ms	2.0	0	2,270	30.0 s
Sustained decode	64K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	64K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	64K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	64K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	64K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	64K	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	1	Valid	38.4	38.6	844 ms	25.9 ms	1.0	0	1,151	30.0 s
Sustained decode	128K	2	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	128K	128	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	1	Valid	31.8	32.0	1,008 ms	31.3 ms	1.0	0	954	30.0 s
Sustained decode	176K	2	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	4	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	8	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	16	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	32	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	64	Over capacity	—	—	—	—	—	—	—	—
Sustained decode	176K	128	Over capacity	—	—	—	—	—	—	—	—

Over capacity: not run because this many requests at this context would exceed the server's KV cache.