# LLM-Benchmark Konfiguration # Modell : Devstral-Small-2-24B-Instruct-2512 # Engine : vLLM # Run-ID : run-20260804-052132-3e1a73 # GPU : NVIDIA GeForce RTX 3090 Ti # CPU : AMD Ryzen 5 5600X 6-Core Processor # RAM : 30 GB bench@llm-benchmark:~$ /opt/vllm-gemma/venv/bin/python /opt/vllm-gemma/venv/bin/vllm serve /home/godcore/models/devstral24-awq \ --served-model-name Devstral-Small-2-24B-Instruct-2512 \ --host 192.168.41.116 \ --port 8000 \ --dtype half \ --max-model-len 32768 \ --gpu-memory-utilization 0.92 \ --enforce-eager \ --limit-mm-per-prompt '{"image":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.	vllm
Modellalias?Der Name, unter dem das Modell ueber die API angesprochen wird. Genau dieser Wert muss im Request-Feld 'model' stehen.	Devstral-Small-2-24B-Instruct-2512
Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann.	32768
Alias?Anzeigename des Modells nach aussen (served model name), unabhaengig vom Dateinamen.	Devstral-Small-2-24B-Instruct-2512
Dtype?Zahlenformat der Modellgewichte bei der Berechnung (z.B. auto, float16, bfloat16). 'auto' waehlt automatisch das vom Modell empfohlene Format.	half
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.	32768
GPU-Speicher?Anteil des GPU-Speichers (0 bis 1), den vLLM belegen darf. 0.92 = 92 %. Hoeher = mehr Platz fuer den KV-Cache (mehr/laengere parallele Anfragen), aber groesseres Risiko fuer 'Out of Memory'.	0.92
Enforce-Eager?Schaltet die optimierte Graph-Ausfuehrung (CUDA-/HIP-Graphs) AB und rechnet Schritt fuer Schritt. Startet schneller und spart etwas VRAM, ist im laufenden Betrieb aber meist langsamer als mit Graphs.	aktiv
limit-mm-per-prompt?Obergrenze fuer multimodale Eingaben pro Prompt (z.B. Anzahl Bilder). Nur fuer multimodale Modelle relevant.	{"image":0}

All benchmarks of this model To leaderboard

Model comparison

Devstral-Small-2-24B-Instruct-2512 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 RTX PRO 6000 Blackwell Workstation Edition 725,6 tok/sNVIDIA GeForce RTX 5090 620,3 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 569,7 tok/s★ NVIDIA GeForce RTX 3090 Ti 355,4 tok/s this runNVIDIA GeForce RTX 5070 Ti 108,5 tok/s

CPUby processor

AMD Ryzen 9 9950X 16-Core Processor 725,6 tok/sAMD Ryzen 7 5800X3D 8-Core Processor 620,3 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 569,7 tok/s★ AMD Ryzen 5 5600X 6-Core Processor 355,4 tok/s this runAMD Ryzen 9 8945HX with Radeon Graphics 331,6 tok/sAMD Ryzen Threadripper PRO 5975WX 32-Cores 108,5 tok/s

MBby mainboard

ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 725,6 tok/sASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 620,3 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 569,7 tok/s★ ASUSTeK COMPUTER INC. PRIME A520M-K 355,4 tok/s this runMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 331,6 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 108,5 tok/s

ENGby engine

★ vLLM 725,6 tok/s this runllama.cpp 645,1 tok/sunbekannt 247,7 tok/s

DRVby driver

unbekannt 725,6 tok/s

💰 Economics

Economics of this run

Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (5× 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 25 W

⚡ TDP 450 W

☁️ External LLM (API) show

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

Electricity0.30 EUR/kWh

Avg power (incl. idle)450 W estimated (TDP)GPU 450 W full load

Avg cost / hourEUR 0.14

Electricity / 1M tokensEUR 0.21

Token / kWh1.46M

Acquisition (system)EUR 1,149 partial priceGPU EUR 999 · PSU EUR 150

Electricity (2 years)–

TCO (2 years)EUR 3,514

Output tokens (2 years)11.50B

☁️ 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 (25 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

	Devstral-Small-2-24B-Instruct-2512NVIDIA GeForce RTX 3090 Ti	Devstral-Small-2-24B-Instruct-2512NVIDIA RTX PRO 6000 Blackwell Workstation Edition	Devstral-Small-2-24B-Instruct-2512NVIDIA RTX PRO 6000 Blackwell Workstation Edition	Devstral-Small-2-24B-Instruct-2512NVIDIA GeForce RTX 5090
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 GeForce RTX 3090 Ti · this run 182 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 423 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 358 tok/s NVIDIA GeForce RTX 5090 340 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

Marcel Sommer

@marcelsommer