ServingCard logoServingCard

qwen3.5-27b-awq

qwen3.5-27b-awq serving configurations from the ServingCard registry.

View on GitHub
Note: This is a dense general-purpose model. PawBench scores measure coding agent performance — this model may excel at different tasks.
Parameters27B (dense)
ArchitectureQwen2ForCausalLM
LicenseApache-2.0
Configs1 variant
Observations2
HuggingFace View on HuggingFace →

Benchmark Observations

mitkoxclaim
Source

Throughput

223.9 tok/s

Latency

GPU

nvidia-gb10

vllm>=0.18.0rc1Unknownturboquant-3.5-triton-native
zenprocessverified
Source

Throughput

27.9 tok/s

Latency

10267ms TTFT

Quality

0.693

GPU

nvidia-gb10

vllm>=0.18.0rc1Unknownturboquant-3.5-triton-native

Fork This Config

Create your own optimized configuration based on these community-verified settings.

Open Configurator

Discussion

Debate benchmarks, share quirks, report what works and what doesn't on your hardware.