Created bymario-alka.dePowered bynoob2claw.detricoma.de
Contributed byMario AlkaPoolside

Laguna-S-2.1-INT4

Performance benchmark · measured on 09.08.2026 11:35

Benchmark-IDrun-20260809-131004-cd69d2
Timebench 3 - Kombi (Prefill + Generation)MoE117.6BRuntime: llama.cppQuantisierung: INT4
Generation9,12tok/s
Prefill157,03tok/s
Time to First Token17.485,50ms
Total duration259,57s
Concurrency1parallel
Ranking in the field
1622of 1927 systems

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

This run is better than 16 % of all comparable systems.
Generation 9,1 tok/s
-88 % vs Ø 77,5
Prefill 157,0 tok/s
-94 % vs Ø 2.844,6
Time to First Token 17.486 ms
-26 % vs Ø 23.755
Distribution in the field0 – 471 tok/s
Ø 78 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

How does this benchmark compare on other GPUs?

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

Configuration

benchmark-konfiguration — run-20260809-131004-cd69d2
# LLM-Benchmark Konfiguration # Modell : Laguna-S-2.1-INT4 # Engine : llama.cpp # Run-ID : run-20260809-131004-cd69d2 # GPU : NVIDIA GeForce RTX 3090 Ti # CPU : AMD Ryzen 9 8945HX with Radeon Graphics # RAM : 92 GB bench@llm-benchmark:~$ /root/llama.cpp/build/bin/llama-server \ -m /root/.cache/huggingface/hub/models--unsloth--Laguna-S-2.1-GGUF/snapshots/750f92f90cf54159c4d7a610cb7b3e74498e75c6/UD-Q4_K_M/Laguna-S-2.1-UD-Q4_K_M-00001-of-00003.gguf \ --alias Laguna-S-2.1-INT4 \ --host 0.0.0.0 \ --port 8000 \ -ngl 12 \ -c 16384 \ -np 4 \ --jinja
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.Laguna-S-2.1-INT4
Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann.16384
Modellpfad?Pfad zur GGUF-Modelldatei, die geladen und ausgeliefert wird./root/.cache/huggingface/hub/models--unsloth--Laguna-S-2.1-GGUF/snapshots/750f92f90cf54159c4d7a610cb7b3e74498e75c6/UD-Q4_K_M/Laguna-S-2.1-UD-Q4_K_M-00001-of-00003.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.12
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.16384
np4
execution_typelocal

All benchmarks of this model To leaderboard

Anzeige
Model comparison

Laguna-S-2.1-INT4 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

1.0097575052520,001.3272.6553.982Prefill (tok/s)Generation (tok/s)NVIDIA RTX PRO 6000 Blackwell Workstation Edition - 781,4 tok/s Generation, 3.226 tok/s Prefill, TTFT 7.860 ms (3 Laufe)NVIDIA RTX PRO 6000 B...NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition - 777,7 tok/s Generation, 3.235 tok/s Prefill, TTFT 9.787 ms (15 Laufe)NVIDIA RTX PRO 6000 B...NVIDIA GeForce RTX 3090 Ti - 32,8 tok/s Generation, 164 tok/s Prefill, TTFT 132.906 ms (3 Laufe) | DIESER LAUF★ NVIDIA GeForce RTX 30...
NVIDIA RTX PRO 6000 Blackwell Workstation Edition 781,4 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 777,7 tok/s★ NVIDIA GeForce RTX 3090 Ti 32,8 tok/s this run

CPUby processor

1.0097575052520,001.3272.6553.982Prefill (tok/s)Generation (tok/s)AMD Ryzen 9 9950X 16-Core Processor - 781,4 tok/s Generation, 3.226 tok/s Prefill, TTFT 7.860 ms (3 Laufe)AMD Ryzen 9 9950X 16-...AMD Ryzen Threadripper PRO 9965WX 24-Cores - 777,7 tok/s Generation, 3.235 tok/s Prefill, TTFT 9.787 ms (15 Laufe)AMD Ryzen Threadrippe...AMD Ryzen 9 8945HX with Radeon Graphics - 32,8 tok/s Generation, 164 tok/s Prefill, TTFT 132.906 ms (3 Laufe) | DIESER LAUF★ AMD Ryzen 9 8945HX wi...
AMD Ryzen 9 9950X 16-Core Processor 781,4 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 777,7 tok/s★ AMD Ryzen 9 8945HX with Radeon Graphics 32,8 tok/s this run

MBby mainboard

1.0097575052520,001.3272.6553.982Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI - 781,4 tok/s Generation, 3.226 tok/s Prefill, TTFT 7.860 ms (3 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 777,7 tok/s Generation, 3.235 tok/s Prefill, TTFT 9.787 ms (15 Laufe)ASUSTeK COMPUTER INC....Meigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) - 32,8 tok/s Generation, 164 tok/s Prefill, TTFT 132.906 ms (3 Laufe) | DIESER LAUF★ Meigao Innovation Tec...
ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 781,4 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 777,7 tok/s★ Meigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 32,8 tok/s this run

ENGby engine

8608217817427032.6272.7392.8512.963Prefill (tok/s)Generation (tok/s)llama.cpp - 781,4 tok/s Generation, 2.795 tok/s Prefill, TTFT 27.100 ms (21 Laufe) | DIESER LAUF★ llama.cpp
★ llama.cpp 781,4 tok/s this run

DRVby driver

8608217817427032.6272.7392.8512.963Prefill (tok/s)Generation (tok/s)unbekannt - 781,4 tok/s Generation, 2.795 tok/s Prefill, TTFT 27.100 ms (21 Laufe)unbekannt
unbekannt 781,4 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 URLReset
⚡ Electricity price EUR/kWh
⚙️ System utilization 100 %
🖥️ Acquisition EUR
🔌 Idle 50 W
⚡ TDP 477 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)477 W estimated (TDP)GPU 450 + CPU 17 + Board 10 W full load
Avg cost / hourEUR 0.14
Electricity / 1M tokensEUR 4.36
Token / kWh68.79K
Acquisition (system)EUR 2,986 partial priceGPU EUR 999 · CPU EUR 549 · RAM EUR 1,288 · PSU EUR 150
Electricity (2 years)–
TCO (2 years)EUR 5,494
Output tokens (2 years)575.22M
☁️ 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 (50 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-S-2.1-INT4NVIDIA GeForce RTX 3090 TiLaguna-S-2.1-INT4NVIDIA RTX PRO 6000 Blackwell Workstation EditionLaguna-S-2.1-INT43x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionLaguna-S-2.1-INT43x NVIDIA RTX PRO 6000 Blackwell Max-Q 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)––––

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.