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

Qwen-AgentWorld-35B-A3B

Performance benchmark · measured on 29.07.2026 04:39

Benchmark-IDrun-20260729-060802-e464c9
Timebench 3 - Kombi (Prefill + Generation)MoE35BRuntime: llama.cppQuantisierung: UD-Q4_K_M
Generation82,00tok/s
Prefill450,79tok/s
Time to First Token34.178,50ms
Total duration265,10s
Concurrency5parallel
Ranking in the field
248of 354 systems

Performance benchmark · Primary metric: Generation-Speed (tok/s) · 5× concurrent

This run is better than 30 % of all comparable systems.
Generation 82,0 tok/s
-78 % vs Ø 370,9
Prefill 450,8 tok/s
-92 % vs Ø 5.422,0
Time to First Token 34.179 ms
+44 % vs Ø 23.732
Distribution in the field0 – 1.349 tok/s
Ø 370 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

gemma-4-E2B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181606-0adf6d
1.349,3 tok/s
gemma-4-E2B-itNVIDIA GeForce RTX 5090 · run-20260728-184455-5b4937
1.301,9 tok/s
gemma-4-E2B-it3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-184456-4111a9
1.195,7 tok/s
NVIDIA-Nemotron-3-Nano-4BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181607-30d864
1.188,7 tok/s
NVIDIA-Nemotron-3-Nano-4BNVIDIA GeForce RTX 5090 · run-20260728-194135-5432cd
1.164,8 tok/s
gpt-oss-20bNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260728-140954-9e81f9
1.136,3 tok/s
gpt-oss-20bNVIDIA GeForce RTX 5090 · run-20260729-032121-058a31
1.113,5 tok/s
NVIDIA-Nemotron-3-Nano-4B3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-836a91
1.090,6 tok/s
NVIDIA-Nemotron-3-Nano-4B3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260727-181606-8b381e
1.077,4 tok/s
NVIDIA-Nemotron-3-Nano-4B3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-b8a818
1.071,0 tok/s
NVIDIA-Nemotron-3-Nano-4B3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-038e92
1.051,1 tok/s
gpt-oss-20b3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260729-032121-ffb18a
1.015,5 tok/s
NVIDIA-Nemotron-3-Nano-30B-A3BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260729-032121-305ad6
1.007,8 tok/s
Nemotron-Cascade-2-30B-A3BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260729-032120-4b8514
1.007,3 tok/s
Qwen-AgentWorld-35B-A3B this runNVIDIA GeForce RTX 5070 Ti · run-20260729-060802-e464c9
82,0 tok/s

How does this benchmark compare on other GPUs?

Same model on different hardware · 5× concurrent · Generation (tok/s)

Hardware

GPU: NVIDIA GeForce RTX 5070 Ti · 16 GB VRAM
CPU: AMD Ryzen Threadripper PRO 5975WX 32-Cores
RAM: 247 GB
Mainboard: ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI

Setup

Runtime: llama.cpp
Quantization: UD-Q4_K_M
Model: Qwen-AgentWorld-35B-A3B

Configuration

benchmark-konfiguration — run-20260729-060802-e464c9
# LLM-Benchmark Konfiguration # Modell : Qwen-AgentWorld-35B-A3B # Engine : llama.cpp # Run-ID : run-20260729-060802-e464c9 # GPU : NVIDIA GeForce RTX 5070 Ti # CPU : AMD Ryzen Threadripper PRO 5975WX 32-Cores # RAM : 247 GB bench@llm-benchmark:~$ /home/godcore/llama.cpp/build/bin/llama-server \ -m /home/godcore/.cache/huggingface/hub/models--unsloth--Qwen-AgentWorld-35B-A3B-GGUF/snapshots/3a305abf5cfd119ee999dfe929c433746edd8d63/Qwen-AgentWorld-35B-A3B-UD-Q4_K_M.gguf \ --alias Qwen-AgentWorld-35B-A3B \ --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.Qwen-AgentWorld-35B-A3B
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./home/godcore/.cache/huggingface/hub/models--unsloth--Qwen-AgentWorld-35B-A3B-GGUF/snapshots/3a305abf5cfd119ee999dfe929c433746edd8d63/Qwen-AgentWorld-35B-A3B-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.12
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

Qwen-AgentWorld-35B-A3B 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

1.8981.4239494740,002.8165.6328.447Prefill (tok/s)Generation (tok/s)NVIDIA RTX PRO 6000 Blackwell Workstation Edition - 1.481,2 tok/s Generation, 6.880 tok/s Prefill, TTFT 3.627 ms (3 Laufe)NVIDIA RTX PRO 6000 B...AMD Radeon AI PRO R9700 - 206,1 tok/s Generation, 3.156 tok/s Prefill, TTFT 4.980 ms (2 Laufe)AMD Radeon AI PRO R97...NVIDIA GeForce RTX 5070 Ti - 140,1 tok/s Generation, 468 tok/s Prefill, TTFT 43.146 ms (3 Laufe) | DIESER LAUF★ NVIDIA GeForce RTX 50...
NVIDIA RTX PRO 6000 Blackwell Workstation Edition 1.481,2 tok/sAMD Radeon AI PRO R9700 206,1 tok/s★ NVIDIA GeForce RTX 5070 Ti 140,1 tok/s this run

CPUby processor

1.8981.4239494740,002.8165.6328.447Prefill (tok/s)Generation (tok/s)AMD Ryzen 9 9950X 16-Core Processor - 1.481,2 tok/s Generation, 6.880 tok/s Prefill, TTFT 3.627 ms (3 Laufe)AMD Ryzen 9 9950X 16-...AMD Ryzen Threadripper PRO 7955WX 16-Cores - 206,1 tok/s Generation, 3.156 tok/s Prefill, TTFT 4.980 ms (2 Laufe)AMD Ryzen Threadrippe...AMD Ryzen Threadripper PRO 5975WX 32-Cores - 140,1 tok/s Generation, 468 tok/s Prefill, TTFT 43.146 ms (3 Laufe) | DIESER LAUF★ AMD Ryzen Threadrippe...
AMD Ryzen 9 9950X 16-Core Processor 1.481,2 tok/sAMD Ryzen Threadripper PRO 7955WX 16-Cores 206,1 tok/s★ AMD Ryzen Threadripper PRO 5975WX 32-Cores 140,1 tok/s this run

MBby mainboard

1.8981.4239494740,002.8165.6328.447Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI - 1.481,2 tok/s Generation, 6.880 tok/s Prefill, TTFT 3.627 ms (3 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 206,1 tok/s Generation, 3.156 tok/s Prefill, TTFT 4.980 ms (2 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI - 140,1 tok/s Generation, 468 tok/s Prefill, TTFT 43.146 ms (3 Laufe) | DIESER LAUF★ ASUSTeK COMPUTER INC....
ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 1.481,2 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 206,1 tok/s★ ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 140,1 tok/s this run

ENGby engine

1.6291.5551.4811.4071.3333.3323.4743.6163.758Prefill (tok/s)Generation (tok/s)llama.cpp - 1.481,2 tok/s Generation, 3.545 tok/s Prefill, TTFT 18.785 ms (8 Laufe) | DIESER LAUF★ llama.cpp
★ llama.cpp 1.481,2 tok/s this run

DRVby driver

1.6291.5551.4811.4071.3333.3323.4743.6163.758Prefill (tok/s)Generation (tok/s)unbekannt - 1.481,2 tok/s Generation, 3.545 tok/s Prefill, TTFT 18.785 ms (8 Laufe)unbekannt
unbekannt 1.481,2 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 URLReset
⚡ Electricity price EUR/kWh
⚙️ System utilization 100 %
🖥️ Acquisition EUR
🔌 Idle 10 W
⚡ TDP 310 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)310 W estimated (TDP)GPU 300 + Board 10 W full load
Avg cost / hourEUR 0.093
Electricity / 1M tokensEUR 0.32
Token / kWh952.26K
Acquisition (system)EUR 2,126 missingRAM EUR 1,976 · PSU EUR 150
Electricity (2 years)
TCO (2 years)EUR 3,755
Output tokens (2 years)5.17B
☁️ 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 (10 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

Qwen-AgentWorld-35B-A3BNVIDIA GeForce RTX 5070 TiQwen-AgentWorld-35B-A3BNVIDIA RTX PRO 6000 Blackwell Workstation EditionQwen-AgentWorld-35B-A3B3x AMD Radeon AI PRO R9700
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.