Created bymario-alka.dePowered bygodcore.denoob2claw.detricoma.de
DEEN
Sign in Sign up
Beta
🎮
Graphics card

NVIDIA GeForce RTX 4090

NVIDIA24 GB VRAM
17Benchmarks
5Models
17Performance
0Harness
450 WMax power
Best generation
503,3 tok/s
Qwen3-0.6B
∅ Generation
184,7 tok/s
Schnitt aus 17 Läufen
Best prefill
112.346 tok/s
Prompt processing
Best TTFT
121 ms
Time to First Token
Best harness rate
–%
noch kein Harness
🏆
Qwen3-0.6B läuft am schnellsten auf dieser GPU · 503,3 tok/s
Bester Prefill: 112.346 tok/s · 5 Modelle getestet

Top models

Bester gemessener Durchsatz je Modell auf NVIDIA GeForce RTX 4090 – umschaltbar nach Generation, Prefill oder Kombiwert.

Qwen3-0.6B0.6B
503,3 tok/s
Meta-Llama-3.1-8B-Instruct8B
489,7 tok/s
Qwen3-Coder-30B-A3B-Instruct30B
400,4 tok/s
Devstral-Small-2-24B-Instruct-251224B
265,9 tok/s
Ministral-3-14B-Reasoning-251214B
157,9 tok/s
Anzeige
Performance profile

Durchsatz auf NVIDIA GeForce RTX 4090

Jede Blase steht für ein Modell, eine Engine bzw. eine CPU – Position: Prompt-Verarbeitung (X) × Ausgabe (Y), Blasengröße: Anzahl Messläufe.

MODnach Modell

62347733118538,4029.25558.50987.764Prefill (tok/s)Generation (tok/s)Qwen3-0.6B - 503,3 tok/s Generation, 71.575 tok/s Prefill, TTFT 174 ms (3 Laufe)Qwen3-0.6BMeta-Llama-3.1-8B-Instruct - 489,7 tok/s Generation, 14.571 tok/s Prefill, TTFT 1.424 ms (3 Laufe)Meta-Llama-3.1-8B-Ins...Qwen3-Coder-30B-A3B-Instruct - 400,4 tok/s Generation, 5.750 tok/s Prefill, TTFT 3.056 ms (5 Laufe)Qwen3-Coder-30B-A3B-I...Devstral-Small-2-24B-Instruct-2512 - 265,9 tok/s Generation, 5.497 tok/s Prefill, TTFT 1.725 ms (4 Laufe)Devstral-Small-2-24B-...Ministral-3-14B-Reasoning-2512 - 157,9 tok/s Generation, 6.785 tok/s Prefill, TTFT 954 ms (2 Laufe)Ministral-3-14B-Reaso...
Qwen3-0.6B 503,3 tok/sMeta-Llama-3.1-8B-Instruct 489,7 tok/sQwen3-Coder-30B-A3B-Instruct 400,4 tok/sDevstral-Small-2-24B-Instruct-2512 265,9 tok/sMinistral-3-14B-Reasoning-2512 157,9 tok/s

ENGnach Engine

574513452391330017.45334.90552.358Prefill (tok/s)Generation (tok/s)vLLM - 503,3 tok/s Generation, 43.073 tok/s Prefill, TTFT 799 ms (6 Laufe)vLLMllama.cpp - 400,4 tok/s Generation, 5.846 tok/s Prefill, TTFT 2.190 ms (11 Laufe)llama.cpp
vLLM 503,3 tok/sllama.cpp 400,4 tok/s

CPUnach Prozessor

55452850347845317.84618.60519.36520.124Prefill (tok/s)Generation (tok/s)AMD Ryzen 3 3100 4-Core Processor - 503,3 tok/s Generation, 18.985 tok/s Prefill, TTFT 1.699 ms (17 Laufe)AMD Ryzen 3 3100 4-Co...
AMD Ryzen 3 3100 4-Core Processor 503,3 tok/s
Throughput & latency

Performancebenchmark

17 veroeffentlichte Performance-Läufe auf NVIDIA GeForce RTX 4090.

In the leaderboard →
Metric:
Size
#Model / MakerMetricsParallelGPU / CPU / RAMRuntime
1Qwen3-0.6B0.6BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
503,33 tok/s TG
Prefill 112.346 · TTFT 242 ms
10×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMvLLMopenclaw_cliFP8Details →
2Meta-Llama-3.1-8B-Instruct8BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
489,72 tok/s TG
Prefill 21.352 · TTFT 2.183 ms
10×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMvLLMopenclaw_cliAWQDetails →
3Qwen3-Coder-30B-A3B-Instruct30BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
400,42 tok/s TG
Prefill 12.694 · TTFT 1.086 ms
5×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
4Devstral-Small-2-24B-Instruct-251224BMistral AI Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
265,94 tok/s TG
Prefill 8.432 · TTFT 3.619 ms
10×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
5Qwen3-0.6B0.6BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
258,23 tok/s TG
Prefill 81.661 · TTFT 159 ms
5×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMvLLMopenclaw_cliFP8Details →
6Meta-Llama-3.1-8B-Instruct8BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
254,69 tok/s TG
Prefill 14.644 · TTFT 1.771 ms
5×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMvLLMopenclaw_cliAWQDetails →
7Qwen3-Coder-30B-A3B-Instruct30BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
198,88 tok/s TG
Prefill 7.188 · TTFT 339 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
8Devstral-Small-2-24B-Instruct-251224BMistral AI Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
186,91 tok/s TG
Prefill 6.423 · TTFT 2.079 ms
5×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
9Ministral-3-14B-Reasoning-251214BMistral AI Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
157,87 tok/s TG
Prefill 8.091 · TTFT 1.495 ms
5×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ8_0Details →
10Qwen3-Coder-30B-A3B-Instruct30BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
68,82 tok/s TG
Prefill 3.989 · TTFT 7.885 ms
10×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
11Qwen3-Coder-30B-A3B-Instruct30BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
58,29 tok/s TG
Prefill 3.005 · TTFT 4.672 ms
5×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
12Meta-Llama-3.1-8B-Instruct8BMeta Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
53,50 tok/s TG
Prefill 7.717 · TTFT 319 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMvLLMopenclaw_cliAWQDetails →
13Qwen3-0.6B0.6BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
52,80 tok/s TG
Prefill 20.717 · TTFT 121 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMvLLMopenclaw_cliFP8Details →
14Ministral-3-14B-Reasoning-251214BMistral AI Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
50,31 tok/s TG
Prefill 5.478 · TTFT 413 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ8_0Details →
15Devstral-Small-2-24B-Instruct-251224BMistral AI Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
49,20 tok/s TG
Prefill 3.616 · TTFT 593 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
16Devstral-Small-2-24B-Instruct-251224BMistral AI Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
49,02 tok/s TG
Prefill 3.516 · TTFT 610 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
17Qwen3-Coder-30B-A3B-Instruct30BQwen (Alibaba) Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
42,25 tok/s TG
Prefill 1.875 · TTFT 1.298 ms
1×NVIDIA GeForce RTX 4090AMD Ryzen 3 3100 4-Core Processor · 61 GB RAMllama.cppopenclaw_cliQ4_K_MDetails →
Agent & chat rating

Harnessbenchmark

0 veroeffentlichte Harness-Läufe auf NVIDIA GeForce RTX 4090.

In the leaderboard →
Noch keine Harnessbenchmarks auf dieser Hardware.

← All hardware

Technische Daten

Technische Daten – NVIDIA GeForce RTX 4090

HerstellerNVIDIA
ArchitekturAda Lovelace
VRAM24 GB
Tensor-Recheneinheiten512
Speicherbandbreite1.008 GB/s
TDP450 W
PCIe4.0
Release10/2022
Best generation503,3 tok/s
Best prefill112.346 tok/s
Models5
Messläufe (Performance)17