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

DeepSeek-V4-Flash

Performance benchmark · measured on 17.08.2026 10:59

Benchmark-IDrun-20260817-120946-8a1de9
Timebench 3 - Kombi (Prefill + Generation)MoE290.9BRuntime: llama.cppQuantisierung: UD-Q4_K_XL
Configuration note: Es sind 4 Grafikkarten installiert, aber nur 3 wurden fuer diesen Lauf genutzt. Mit allen 4 GPUs koennte der Durchsatz deutlich hoeher liegen.
Generation14,49tok/s
Prefill162,20tok/s
Time to First Token65.453,00ms
Total duration77,14s
Concurrency5parallel
Ranking in the field
166of 203 systems

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

This run is better than 18 % of all comparable systems.
Generation 14,5 tok/s
-82 % vs Ø 79,5
Prefill 162,2 tok/s
-91 % vs Ø 1.742,0
Time to First Token 65.453 ms
-19 % vs Ø 81.026
Distribution in the field0 – 382 tok/s
Ø 80 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

Hardware

GPU: 4x AMD Radeon AI PRO R9700 · 32 GB VRAM · only 3 of 4 used
CPU: AMD Ryzen Threadripper PRO 7955WX 16-Cores
RAM: 184 GB
Mainboard: ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE

Setup

Runtime: llama.cpp
Quantization: UD-Q4_K_XL
Model: DeepSeek-V4-Flash

Configuration

benchmark-konfiguration — run-20260817-120946-8a1de9
# LLM-Benchmark Konfiguration # Modell : DeepSeek-V4-Flash # Engine : llama.cpp # Run-ID : run-20260817-120946-8a1de9 # GPU : 4x AMD Radeon AI PRO R9700 # CPU : AMD Ryzen Threadripper PRO 7955WX 16-Cores # RAM : 184 GB bench@llm-benchmark:~$ llama-server \ --model DeepSeek-V4-Flash
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.DeepSeek-V4-Flash

All benchmarks of this model To leaderboard

Anzeige
Model comparison

DeepSeek-V4-Flash 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

22,721,620,619,618,5157164170177Prefill (tok/s)Generation (tok/s)AMD Radeon AI PRO R9700 - 20,6 tok/s Generation, 167 tok/s Prefill, TTFT 65.962 ms (15 Laufe) | DIESER LAUF★ AMD Radeon AI PRO R97...
★ AMD Radeon AI PRO R9700 20,6 tok/s this run

CPUby processor

22,721,620,619,618,5157164170177Prefill (tok/s)Generation (tok/s)AMD Ryzen Threadripper PRO 7955WX 16-Cores - 20,6 tok/s Generation, 167 tok/s Prefill, TTFT 65.962 ms (15 Laufe) | DIESER LAUF★ AMD Ryzen Threadrippe...
★ AMD Ryzen Threadripper PRO 7955WX 16-Cores 20,6 tok/s this run

MBby mainboard

22,721,620,619,618,5157164170177Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 20,6 tok/s Generation, 167 tok/s Prefill, TTFT 65.962 ms (15 Laufe) | DIESER LAUF★ ASUSTeK COMPUTER INC....
★ ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 20,6 tok/s this run

ENGby engine

22,721,620,619,618,5157164170177Prefill (tok/s)Generation (tok/s)llama.cpp - 20,6 tok/s Generation, 167 tok/s Prefill, TTFT 65.962 ms (15 Laufe) | DIESER LAUF★ llama.cpp
★ llama.cpp 20,6 tok/s this run

DRVby driver

22,721,620,619,618,5157164170177Prefill (tok/s)Generation (tok/s)unbekannt - 20,6 tok/s Generation, 167 tok/s Prefill, TTFT 65.962 ms (15 Laufe)unbekannt
unbekannt 20,6 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 URLReset
⚡ Electricity price EUR/kWh
⚙️ System utilization 100 %
🖥️ Acquisition EUR
🔌 Idle 145 W
⚡ TDP 1271 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)1,271 W estimated (TDP)GPU 1,200 + CPU 46 + Board 25 W full load
Avg cost / hourEUR 0.38
Electricity / 1M tokensEUR 7.31
Token / kWh41.04K
Acquisition (system)EUR 11,114 full priceGPU EUR 5,600 · CPU EUR 1,399 · Board EUR 1,299 · RAM EUR 2,576 · PSU EUR 240
Electricity (2 years)
TCO (2 years)EUR 17,794
Output tokens (2 years)913.91M
☁️ 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 (145 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

DeepSeek-V4-Flash4x AMD Radeon AI PRO R9700DeepSeek-V4-Flash4x AMD Radeon AI PRO R9700DeepSeek-V4-Flash4x AMD Radeon AI PRO R9700DeepSeek-V4-Flash4x 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.