Created bymario-alka.dePowered bygodcore.denoob2claw.detricoma.de
Contributed byMario AlkaDeepReinforce

Ornith-1.0-9B

Performance benchmark · measured on 30.07.2026 04:37

Benchmark-IDrun-20260730-064527-429cd1
Timebench 3 - Kombi (Prefill + Generation)Dense9BRuntime: llama.cppQuantisierung: Q4_K_M
Generation1.505,61tok/s
Prefill8.189,59tok/s
Time to First Token7.530,50ms
Total duration44,91s
Concurrency10parallel
Ranking in the field
20of 48 systems

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

This run is better than 60 % of all comparable systems.
Generation 1.505,6 tok/s
+48 % vs Ø 1.018,7
Prefill 8.189,6 tok/s
-11 % vs Ø 9.196,5
Time to First Token 7.531 ms
-79 % vs Ø 36.381
Distribution in the field1 – 2.491 tok/s
Ø 1.019 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: NVIDIA GeForce RTX 5090 · 32 GB VRAM
CPU: AMD Ryzen 7 5800X3D 8-Core Processor
RAM: 126 GB
Mainboard: ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING

Setup

Runtime: llama.cpp
Quantization: Q4_K_M
Model: Ornith-1.0-9B

Configuration

benchmark-konfiguration — run-20260730-064527-429cd1
# LLM-Benchmark Konfiguration # Modell : Ornith-1.0-9B # Engine : llama.cpp # Run-ID : run-20260730-064527-429cd1 # GPU : NVIDIA GeForce RTX 5090 # CPU : AMD Ryzen 7 5800X3D 8-Core Processor # RAM : 126 GB bench@llm-benchmark:~$ /root/llama.cpp/build/bin/llama-server \ -m /root/.cache/huggingface/hub/models--deepreinforce-ai--Ornith-1.0-9B-GGUF/snapshots/3296bc7a404871a72ac3f1903f561459c09b5c17/ornith-1.0-9b-Q4_K_M.gguf \ --alias Ornith-1.0-9B \ --host 0.0.0.0 \ --port 8000 \ -ngl 999 \ -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.Ornith-1.0-9B
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--deepreinforce-ai--Ornith-1.0-9B-GGUF/snapshots/3296bc7a404871a72ac3f1903f561459c09b5c17/ornith-1.0-9b-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
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.16384
np4

All benchmarks of this model To leaderboard

Anzeige
Model comparison

Ornith-1.0-9B 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

1.8521.5371.2229075923.1655.5827.99910.416Prefill (tok/s)Generation (tok/s)NVIDIA RTX PRO 6000 Blackwell Workstation Edition - 1.560,5 tok/s Generation, 9.057 tok/s Prefill, TTFT 3.092 ms (3 Laufe)NVIDIA RTX PRO 6000 B...NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition - 1.349,7 tok/s Generation, 7.752 tok/s Prefill, TTFT 3.646 ms (9 Laufe)NVIDIA RTX PRO 6000 B...NVIDIA GeForce RTX 5070 Ti - 883,4 tok/s Generation, 4.525 tok/s Prefill, TTFT 5.715 ms (3 Laufe)NVIDIA GeForce RTX 50...NVIDIA GeForce RTX 5090 - 1.505,6 tok/s Generation, 6.762 tok/s Prefill, TTFT 3.486 ms (3 Laufe) | DIESER LAUF★ NVIDIA GeForce RTX 50...
NVIDIA RTX PRO 6000 Blackwell Workstation Edition 1.560,5 tok/s★ NVIDIA GeForce RTX 5090 1.505,6 tok/s this runNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.349,7 tok/sNVIDIA GeForce RTX 5070 Ti 883,4 tok/s

CPUby processor

1.8521.5371.2229075923.1655.5827.99910.416Prefill (tok/s)Generation (tok/s)AMD Ryzen 9 9950X 16-Core Processor - 1.560,5 tok/s Generation, 9.057 tok/s Prefill, TTFT 3.092 ms (3 Laufe)AMD Ryzen 9 9950X 16-...AMD Ryzen Threadripper PRO 9965WX 24-Cores - 1.349,7 tok/s Generation, 7.752 tok/s Prefill, TTFT 3.646 ms (9 Laufe)AMD Ryzen Threadrippe...AMD Ryzen Threadripper PRO 5975WX 32-Cores - 883,4 tok/s Generation, 4.525 tok/s Prefill, TTFT 5.715 ms (3 Laufe)AMD Ryzen Threadrippe...AMD Ryzen 7 5800X3D 8-Core Processor - 1.505,6 tok/s Generation, 6.762 tok/s Prefill, TTFT 3.486 ms (3 Laufe) | DIESER LAUF★ AMD Ryzen 7 5800X3D 8...
AMD Ryzen 9 9950X 16-Core Processor 1.560,5 tok/s★ AMD Ryzen 7 5800X3D 8-Core Processor 1.505,6 tok/s this runAMD Ryzen Threadripper PRO 9965WX 24-Cores 1.349,7 tok/sAMD Ryzen Threadripper PRO 5975WX 32-Cores 883,4 tok/s

MBby mainboard

1.8521.5371.2229075923.1655.5827.99910.416Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI - 1.560,5 tok/s Generation, 9.057 tok/s Prefill, TTFT 3.092 ms (3 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 1.349,7 tok/s Generation, 7.752 tok/s Prefill, TTFT 3.646 ms (9 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI - 883,4 tok/s Generation, 4.525 tok/s Prefill, TTFT 5.715 ms (3 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING - 1.505,6 tok/s Generation, 6.762 tok/s Prefill, TTFT 3.486 ms (3 Laufe) | DIESER LAUF★ ASUSTeK COMPUTER INC....
ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 1.560,5 tok/s★ ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.505,6 tok/s this runASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 1.349,7 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 883,4 tok/s

ENGby engine

1.7171.6391.5611.4831.4046.8317.1217.4127.703Prefill (tok/s)Generation (tok/s)llama.cpp - 1.560,5 tok/s Generation, 7.267 tok/s Prefill, TTFT 3.872 ms (18 Laufe) | DIESER LAUF★ llama.cpp
★ llama.cpp 1.560,5 tok/s this run

DRVby driver

1.7171.6391.5611.4831.4046.8317.1217.4127.703Prefill (tok/s)Generation (tok/s)unbekannt - 1.560,5 tok/s Generation, 7.267 tok/s Prefill, TTFT 3.872 ms (18 Laufe)unbekannt
unbekannt 1.560,5 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 72 W
⚡ TDP 622 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)622 W estimated (TDP)GPU 575 + CPU 35 + Board 12 W full load
Avg cost / hourEUR 0.19
Electricity / 1M tokensEUR 0.034
Token / kWh8.72M
Acquisition (system)EUR 4,986 full priceGPU EUR 3,300 · CPU EUR 349 · Board EUR 149 · RAM EUR 1,008 · PSU EUR 180
Electricity (2 years)
TCO (2 years)EUR 8,253
Output tokens (2 years)94.96B
☁️ 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 (72 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

Ornith-1.0-9BNVIDIA GeForce RTX 5090Ornith-1.0-9BNVIDIA RTX PRO 6000 Blackwell Workstation EditionOrnith-1.0-9B3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation EditionOrnith-1.0-9B3x 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.