Qwen3.8-27B · NVFP4
This page aggregates 1 real-world runs of Qwen3.8-27B (NVFP4) on NVIDIA RTX 5090 Laptop with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
77.1 tok/s
Decode
Decode speed
32.7 tok/s
Prefill
Prefill speed
— s
TTFT
Time to first token
— GB
VRAM
VRAM usage
1
Measured runs
1
Independent sources
GitHub
Source platforms
Yesterday
Last verified
Performance
- llama.cpp · NVFP4 (current)Decode 77.1 · Prefill 32.7 ·
Core figures
- Decode (avg)
- 77.1 tok/s
- Prefill (avg)
- 32.7 tok/s
- TTFT (avg)
- — s
- VRAM (avg)
- — GB
- MTP acceptance rate
- —
- TTFB
- — GB
- Power draw
- — W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3.8-27B
- Quantization
- NVFP4
- Framework
- llama.cpp
- Version
- b10915
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA RTX 5090 Laptop
- Nominal VRAM
- 24 GB
- OS
- Windows 11
Get started
Original model resources
Base model resources may not match this quantization. Use the verified deployment weights above when available.
Sources & evidence
1 measured records in total, each traceable to its original source
b10915 · Windows 11
77.1 tok/s
Decode
32.7 tok/s
Prefill
— s
TTFT
0 GB
VRAM
—
MTP
— W
Power draw
self-compiled + MTP | 32.7 tok/s | 77.1 tok/s; self-compiled, MTP disabled | 1867.5 tok/s | 39.3 tok/s