Created bymario-alka.dePowered bygodcore.denoob2claw.detricoma.de
Contributed byMario AlkaQwen (Alibaba)

Qwen3.8-27B

Performance benchmark · measured on 22.09.2026 10:04

Benchmark-IDrun-20260922-081655-96d95a
Timebench 3 - Kombi (Prefill + Generation)Qwen3_5ForConditionalGeneration (multimodal, image-text-to-text)27BRuntime: llama.cppQuantisierung: Q4_K_M
Generation117,36tok/s
Prefill1.455,22tok/s
Time to First Token15.755,00ms
Total duration207,17s
Concurrency10parallel
Ranking in the field
108of 234 systems

Performance benchmark · Primary metric: Generation-Speed (tok/s) · 10× concurrent

This run is better than 54 % of all comparable systems.
Generation 117,4 tok/s
-16 % vs Ø 139,6
Prefill 1.455,2 tok/s
-48 % vs Ø 2.798,0
Time to First Token 15.755 ms
-84 % vs Ø 101.035
Distribution in the field0 – 667 tok/s
Ø 140 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

How does this benchmark compare on other GPUs?

Same model on different hardware · 10× concurrent · Generation (tok/s)

Hardware

GPU: AMD Radeon AI PRO R9700 · 32 GB VRAM
CPU: AMD Ryzen 3 3100 4-Core Processor
RAM: 31 GB
Mainboard: ASUSTeK COMPUTER INC. ROG STRIX X570-F GAMING

Setup

Runtime: llama.cpp
Quantization: Q4_K_M
Model: Qwen3.8-27B

Configuration

benchmark-konfiguration — run-20260922-081655-96d95a
# LLM-Benchmark Konfiguration # Modell : Qwen3.8-27B # Engine : llama.cpp # Run-ID : run-20260922-081655-96d95a # GPU : AMD Radeon AI PRO R9700 # CPU : AMD Ryzen 3 3100 4-Core Processor # RAM : 31 GB bench@llm-benchmark:~$ llama-server \ -m /home/godcore/models/Qwen3.8-27B-UD-Q4_K_M.gguf \ --alias Qwen3.8-27B \ -ngl 999 \ -fa on \ -c 40960 \ -np 10 '(AMD' Radeon AI PRO R9700, ROCm 7.2.4, 'HIP_VISIBLE_DEVICES=0)'
Engine?Die Inferenz-Software, die das Modell ausliefert (z.B. vLLM oder llama.cpp). Sie bestimmt Geschwindigkeit, unterstuetzte Modellformate und welche Parameter ueberhaupt verfuegbar sind.llamacpp
Modellalias?Der Name, unter dem das Modell ueber die API angesprochen wird. Genau dieser Wert muss im Request-Feld 'model' stehen.Qwen3.8-27B
Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann.40960
Modellpfad?Pfad zur GGUF-Modelldatei, die geladen und ausgeliefert wird./home/godcore/models/Qwen3.8-27B-UD-Q4_K_M.gguf
GPU-Layer?Anzahl der auf die GPU ausgelagerten Modell-Layer. Hoeher = mehr VRAM und schneller; der Rest laeuft auf der CPU. 999 = alles auf GPU.999
faon
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.40960
np10

All benchmarks of this model To leaderboard

Anzeige
Model comparison

Qwen3.8-27B on various hardware

All published performance runs of this model – each bubble a variant: position = prefill (X) × generation (Y), bubble size = number of runs. Closer to the top right = faster. ★ Marked gold = this benchmark.

GPUby graphics card

5094183282381471.3992.7014.0025.303Prefill (tok/s)Generation (tok/s)NVIDIA RTX A6000 - 426,5 tok/s Generation, 4.584 tok/s Prefill, TTFT 2.871 ms (15 Laufe)NVIDIA RTX A6000AMD Radeon AI PRO R9700 - 229,4 tok/s Generation, 2.118 tok/s Prefill, TTFT 3.148 ms (39 Laufe) | DIESER LAUF★ AMD Radeon AI PRO R97...
NVIDIA RTX A6000 426,5 tok/s★ AMD Radeon AI PRO R9700 229,4 tok/s this run

CPUby processor

53140127214212,96111.5412.4713.400Prefill (tok/s)Generation (tok/s)AMD Ryzen Threadripper PRO 7955WX 16-Cores - 426,5 tok/s Generation, 2.903 tok/s Prefill, TTFT 2.696 ms (51 Laufe)AMD Ryzen Threadrippe...AMD Ryzen 3 3100 4-Core Processor - 117,4 tok/s Generation, 1.108 tok/s Prefill, TTFT 9.445 ms (3 Laufe) | DIESER LAUF★ AMD Ryzen 3 3100 4-Co...
AMD Ryzen Threadripper PRO 7955WX 16-Cores 426,5 tok/s★ AMD Ryzen 3 3100 4-Core Processor 117,4 tok/s this run

MBby mainboard

53140127214212,96111.5412.4713.400Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 426,5 tok/s Generation, 2.903 tok/s Prefill, TTFT 2.696 ms (51 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. ROG STRIX X570-F GAMING - 117,4 tok/s Generation, 1.108 tok/s Prefill, TTFT 9.445 ms (3 Laufe) | DIESER LAUF★ ASUSTeK COMPUTER INC....
ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 426,5 tok/s★ ASUSTeK COMPUTER INC. ROG STRIX X570-F GAMING 117,4 tok/s this run

ENGby engine

53140127214212,91.0792.2283.3784.528Prefill (tok/s)Generation (tok/s)vLLM - 426,5 tok/s Generation, 3.899 tok/s Prefill, TTFT 3.718 ms (27 Laufe)vLLMllama.cpp - 117,4 tok/s Generation, 1.707 tok/s Prefill, TTFT 2.424 ms (27 Laufe) | DIESER LAUF★ llama.cpp
vLLM 426,5 tok/s★ llama.cpp 117,4 tok/s this run

DRVby driver

4694484274053842.6352.7472.8592.971Prefill (tok/s)Generation (tok/s)unbekannt - 426,5 tok/s Generation, 2.803 tok/s Prefill, TTFT 3.071 ms (54 Laufe)unbekannt
unbekannt 426,5 tok/s
💰 Economics

Economics of this run

Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (10× concurrent). Methodology →

⚙️ ConfigurationAll metrics and charts below follow these settings – based on a 24-month runtime.Save to URLReset
⚡ Electricity price EUR/kWh
⚙️ System utilization 100 %
🖥️ Acquisition EUR
🔌 Idle 20 W
⚡ TDP 300 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)300 W estimated (TDP)GPU 300 W full load
Avg cost / hourEUR 0.090
Electricity / 1M tokensEUR 0.21
Token / kWh1.41M
Acquisition (system)EUR 1,520 partial priceGPU EUR 1,400 · PSU EUR 120
Electricity (2 years)
TCO (2 years)EUR 3,097
Output tokens (2 years)7.40B
☁️ External LLM (API) – comparison
External LLM cost (2 years)
Savings vs. external (2 years)

All values above and the charts below take the configured system utilization into account: at X% the system generates only X% of the time, the rest it idles (20 W). Cost per hour drops (more idle), cost per token rises.

Cost over 2 years – electricity only

Cost over 2 years – incl. acquisition (TCO)

Speed vs. tokens per euro

Euro per 1M tokens

Comparison vs. API – economics per benchmark

Qwen3.8-27BAMD Radeon AI PRO R9700Qwen3.8-27B2x NVIDIA RTX A6000Qwen3.8-27B2x NVIDIA RTX A6000Qwen3.8-27B2x NVIDIA RTX A6000
Electricity cost (24 mo.)
Acquisition cost
Total cost (TCO)
Generated tokens (24 mo.)
Token price via API
Break-even point (days)
Result (savings / extra cost)

Comparison with up to 3 next-best runs of this model at the same concurrency (at least one on different hardware). Power = GPU TDP + CPU (idle + 15 %) + board (estimated), acquisition = full system (GPU + CPU + board + RAM + PSU), prices = stored market prices.

Contributed by

Mario Alka Administrator

@marioalka

Ich bin Unternehmer, Softwareentwickler und KI-Enthusiast. Seit vielen Jahren entwickle ich Unternehmenssoftware und beschäftige mich inzwischen fast täglich mit lokalen LLMs, KI-Agenten und leistungsfähiger KI-Hardware.

Mit LLM-Benchmark.de möchte ich eine Plattform schaffen, auf der Modelle, GPUs und Agenten objektiv und reproduzierbar miteinander verglichen werden.