⚠

Configuration note: Es sind 2 Grafikkarten installiert, aber nur 1 wurde fuer diesen Lauf genutzt. Mit allen 2 GPUs koennte der Durchsatz deutlich hoeher liegen.

Generation94,59tok/s

Prefill7.915,33tok/s

Time to First Token346,50ms

Total duration22,35s

Concurrency1parallel

Ranking in the field

This GPU All GPUs

53of 101 systems

Performance benchmark · Primary metric: Generation-Speed (tok/s) · 1× concurrent

This run is better than 48 % of all comparable systems.

Generation 94,6 tok/s

+3 % vs Ø 92,1

Prefill 7.915,3 tok/s

+74 % vs Ø 4.541,8

Time to First Token 347 ms

-52 % vs Ø 729

Distribution in the field5 – 194 tok/s

Ø 92 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

gemma-4-E2B-itNVIDIA RTX A6000 · run-20260826-072502-338126

194,5 tok/s

Nemotron-3-Nano-30B-A3BNVIDIA RTX A6000 · run-20260826-211812-1bfe3e

180,6 tok/s

Nemotron-Cascade-2-30B-A3BNVIDIA RTX A6000 · run-20260826-211811-bea37a

180,6 tok/s

gpt-oss-20bNVIDIA RTX A6000 · run-20260826-103849-5886ce

175,1 tok/s

Nemotron-3-Nano-4BNVIDIA RTX A6000 · run-20260826-072502-685e56

166,5 tok/s

gpt-oss-20bNVIDIA RTX A6000 · run-20260826-054845-3a8cec

165,8 tok/s

gpt-oss-20bNVIDIA RTX A6000 · run-20260826-103849-b61a63

164,9 tok/s

Qwen3-Coder-30B-A3B-InstructNVIDIA RTX A6000 · run-20260826-054846-d66c35

161,6 tok/s

Qwen3-30B-A3B-Instruct-2507NVIDIA RTX A6000 · run-20260826-054845-6cfbb2

161,0 tok/s

Qwen3-Omni-30B-A3B-ThinkingNVIDIA RTX A6000 · run-20260826-142151-91fc3e

161,0 tok/s

Qwen3-30B-A3B-Thinking-2507NVIDIA RTX A6000 · run-20260826-134908-b256ef

160,7 tok/s

Nemotron-3-Nano-30B-A3BNVIDIA RTX A6000 · run-20260826-134908-e6d3b0

159,7 tok/s

Qwen3-VL-30B-A3B-InstructNVIDIA RTX A6000 · run-20260826-145351-5a8761

157,6 tok/s

Nemotron-Cascade-2-30B-A3BNVIDIA RTX A6000 · run-20260826-062050-97df90

157,0 tok/s

Laguna-XS-2.1 this run2× NVIDIA RTX A6000 · run-20260827-024202-a5f9eb

94,6 tok/s

How does this benchmark compare on other GPUs?

Same model on different hardware · 1× concurrent · Generation (tok/s)

NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260728-140954-cecd39

310,2 tok/s

NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260728-140955-42efc5

309,9 tok/s

NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260728-140955-abb113

309,5 tok/s

NVIDIA GeForce RTX 5090run-20260729-032122-816f0e

308,3 tok/s

NVIDIA GeForce RTX 5090run-20260729-032122-45ddd9

308,2 tok/s

NVIDIA GeForce RTX 5090run-20260729-032123-072f17

307,9 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032123-9ed93e

276,1 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032122-868cb3

275,7 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032124-0504c1

275,5 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032123-56d822

274,3 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032123-9dad71

273,9 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032124-817c48

273,8 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032124-64abce

273,6 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-032122-fb0cea

273,6 tok/s

Hardware

GPU: 2x NVIDIA RTX A6000 · 48 GB VRAM · only 1 of 2 used
CPU: AMD Ryzen Threadripper PRO 7955WX 16-Cores
RAM: 184 GB
Mainboard: ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE

Setup

Runtime: vLLM
Quantization: FP8
Model: Laguna-XS-2.1

Configuration

benchmark-konfiguration — run-20260827-024202-a5f9eb

# LLM-Benchmark Konfiguration # Modell : Laguna-XS-2.1 # Engine : vLLM # Run-ID : run-20260827-024202-a5f9eb # GPU : 2x NVIDIA RTX A6000 # CPU : AMD Ryzen Threadripper PRO 7955WX 16-Cores # RAM : 184 GB bench@llm-benchmark:~$ vllm serve \ --model Laguna-XS-2.1

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.	vllm
Modellalias?Der Name, unter dem das Modell ueber die API angesprochen wird. Genau dieser Wert muss im Request-Feld 'model' stehen.	Laguna-XS-2.1

All benchmarks of this model To leaderboard

Model comparison

Laguna-XS-2.1 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

NVIDIA GeForce RTX 5090 1.765,0 tok/sNVIDIA RTX PRO 6000 Blackwell Workstation Edition 1.757,4 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.629,9 tok/sNVIDIA GeForce RTX 3090 Ti 986,1 tok/s★ NVIDIA RTX A6000 488,9 tok/s this runNVIDIA GeForce RTX 5070 Ti 319,0 tok/sAMD Radeon AI PRO R9700 154,3 tok/s

CPUby processor

AMD Ryzen 7 5800X3D 8-Core Processor 1.765,0 tok/sAMD Ryzen 9 9950X 16-Core Processor 1.757,4 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 1.629,9 tok/sAMD Ryzen 9 8945HX with Radeon Graphics 986,1 tok/s★ AMD Ryzen Threadripper PRO 7955WX 16-Cores 488,9 tok/s this runAMD Ryzen Threadripper PRO 5975WX 32-Cores 319,0 tok/s

MBby mainboard

ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.765,0 tok/sASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 1.757,4 tok/s★ ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 1.629,9 tok/s this runMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 986,1 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 319,0 tok/s

ENGby engine

llama.cpp 1.765,0 tok/s★ vLLM 488,9 tok/s this rununbekannt 26,9 tok/s

DRVby driver

unbekannt 1.765,0 tok/sAMD 7.0.0-27-generic 26,9 tok/s

💰 Economics

Economics of this run

Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (1× concurrent). Methodology →

⚙️ ConfigurationAll metrics and charts below follow these settings – based on a 24-month runtime.Save to URL Reset

⚡ Electricity price EUR/kWh

⚙️ System utilization 100 %

🖥️ Acquisition EUR

🔌 Idle 80 W

⚡ TDP 371 W

☁️ External LLM (API) show

Input €/1M Output €/1M Input per output token

Electricity0.30 EUR/kWh

Avg power (incl. idle)371 W estimated (TDP)GPU 300 + CPU 46 + Board 25 W full load

Avg cost / hourEUR 0.11

Electricity / 1M tokensEUR 0.33

Token / kWh917.85K

Acquisition (system)EUR 8,424 full priceGPU EUR 3,000 · CPU EUR 1,399 · Board EUR 1,299 · RAM EUR 2,576 · PSU EUR 150

Electricity (2 years)–

TCO (2 years)EUR 10,374

Output tokens (2 years)5.97B

☁️ 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 (80 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

	Laguna-XS-2.1NVIDIA RTX A6000	Laguna-XS-2.1NVIDIA RTX PRO 6000 Blackwell Workstation Edition	Laguna-XS-2.1NVIDIA RTX PRO 6000 Blackwell Workstation Edition	Laguna-XS-2.1NVIDIA RTX PRO 6000 Blackwell Workstation Edition
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)	–	–	–	–

NVIDIA RTX A6000 · this run 95 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 310 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 310 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 310 tok/s

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.