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

MiniMax-M2.5

Performance benchmark · measured on 28.07.2026 14:25

Benchmark-IDrun-20260728-143206-f6355b
Timebench 3 - Kombi (Prefill + Generation)MoE230BRuntime: llama.cppQuantisierung: Q4_K_M
Generation6,02tok/s
Prefill65,15tok/s
Time to First Token35.246,50ms
Total duration411,79s
Concurrency1parallel
Ranking in the field
28of 29 systems

Performance benchmark · Primary metric: Generation-Speed (tok/s) · 1× concurrent

This run is better than 4 % of all comparable systems.
Generation 6,0 tok/s
-96 % vs Ø 169,0
Prefill 65,2 tok/s
-99 % vs Ø 5.405,2
Time to First Token 35.247 ms
+1.550 % vs Ø 2.136
Distribution in the field0 – 393 tok/s
Ø 169 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

NVIDIA-Nemotron-3-Nano-4BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181607-bffb11
393,5 tok/s
gemma-4-E2B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181606-6ea089
378,6 tok/s
Laguna-XS-2.1NVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260728-140954-cecd39
310,2 tok/s
Laguna-XS-2.1NVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260728-140955-42efc5
309,9 tok/s
Laguna-XS-2.1NVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260728-140955-abb113
309,5 tok/s
gpt-oss-20bNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260723-194447-636902
268,7 tok/s
gemma-4-E4B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181607-d807ba
246,5 tok/s
Qwen3-Coder-30B-A3B-InstructNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-004605-d79930
237,2 tok/s
Qwen3-30B-A3B-Thinking-2507NVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-004605-ffa4c3
233,9 tok/s
Qwen3-Omni-30B-A3B-ThinkingNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-004611-89d26d
229,0 tok/s
Qwen3-VL-30B-A3B-InstructNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-004607-ae055d
227,9 tok/s
Qwen3-30B-A3B-Instruct-2507NVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-004604-055687
204,9 tok/s
Qwen3.5-35B-A3BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-082959-ff0f63
195,5 tok/s
Qwen3.6-35B-A3BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260724-083000-d5c2ec
195,0 tok/s
MiniMax-M2.5 this runNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260728-143206-f6355b
6,0 tok/s

How does this benchmark compare on other GPUs?

Same model on different hardware · 1× concurrent · Generation (tok/s)

Hardware

GPU: NVIDIA RTX PRO 6000 Blackwell Workstation Edition · 96 GB VRAM
CPU: AMD Ryzen 9 9950X 16-Core Processor
RAM: 92 GB
Mainboard: ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI

Setup

Runtime: llama.cpp
Quantization: Q4_K_M
Model: MiniMax-M2.5

Configuration

benchmark-konfiguration — run-20260728-143206-f6355b
# LLM-Benchmark Konfiguration # Modell : MiniMax-M2.5 # Engine : llama.cpp # Run-ID : run-20260728-143206-f6355b # GPU : NVIDIA RTX PRO 6000 Blackwell Workstation Edition # CPU : AMD Ryzen 9 9950X 16-Core Processor # RAM : 92 GB bench@llm-benchmark:~$ /home/godcore/llama.cpp/build/bin/llama-server \ -m /home/godcore/.cache/huggingface/hub/models--exdysa--MiniMax-M2.5-REAP-172B-A10B-GGUF-Q4_K_M/snapshots/c5518e6b02673458388328830793414ef5bfba5c/MiniMax-M2.5-REAP-172B-A10B-Q4_K_M.gguf \ --alias MiniMax-M2.5 \ --host 0.0.0.0 \ --port 8000 \ -ngl 30 \ -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.MiniMax-M2.5
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--exdysa--MiniMax-M2.5-REAP-172B-A10B-GGUF-Q4_K_M/snapshots/c5518e6b02673458388328830793414ef5bfba5c/MiniMax-M2.5-REAP-172B-A10B-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.30
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

MiniMax-M2.5 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

13410167,233,60,002.0184.0376.055Prefill (tok/s)Generation (tok/s)NVIDIA GB10 (DGX Spark) - 103,5 tok/s Generation, 4.887 tok/s Prefill, TTFT 2.995 ms (3 Laufe)NVIDIA GB10 (DGX Spar...AMD Radeon AI PRO R9700 - 7,0 tok/s Generation, 399 tok/s Prefill, TTFT 41.010 ms (3 Laufe)AMD Radeon AI PRO R97...CPU-only - 0,8 tok/s Generation, 22 tok/s Prefill, TTFT 98.716 ms (1 Lauf)CPU-onlyNVIDIA RTX PRO 6000 Blackwell Workstation Edition - 6,0 tok/s Generation, 65 tok/s Prefill, TTFT 35.247 ms (1 Lauf) | DIESER LAUF★ NVIDIA RTX PRO 6000 B...
NVIDIA GB10 (DGX Spark) 103,5 tok/sAMD Radeon AI PRO R9700 7,0 tok/s★ NVIDIA RTX PRO 6000 Blackwell Workstation Edition 6,0 tok/s this runCPU-only 0,8 tok/s

CPUby processor

13410167,233,60,002.0184.0376.055Prefill (tok/s)Generation (tok/s)NVIDIA Grace - 103,5 tok/s Generation, 4.887 tok/s Prefill, TTFT 2.995 ms (3 Laufe)NVIDIA GraceAMD Ryzen Threadripper PRO 7955WX 16-Cores - 7,0 tok/s Generation, 399 tok/s Prefill, TTFT 41.010 ms (3 Laufe)AMD Ryzen Threadrippe...Intel(R) Xeon(R) CPU E5-4620 v2 @ 2.60GHz - 0,8 tok/s Generation, 22 tok/s Prefill, TTFT 98.716 ms (1 Lauf)Intel(R) Xeon(R) CPU ...AMD Ryzen 9 9950X 16-Core Processor - 6,0 tok/s Generation, 65 tok/s Prefill, TTFT 35.247 ms (1 Lauf) | DIESER LAUF★ AMD Ryzen 9 9950X 16-...
NVIDIA Grace 103,5 tok/sAMD Ryzen Threadripper PRO 7955WX 16-Cores 7,0 tok/s★ AMD Ryzen 9 9950X 16-Core Processor 6,0 tok/s this runIntel(R) Xeon(R) CPU E5-4620 v2 @ 2.60GHz 0,8 tok/s

MBby mainboard

13410167,233,60,002.0184.0376.055Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. GX10 - 103,5 tok/s Generation, 4.887 tok/s Prefill, TTFT 2.995 ms (3 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 7,0 tok/s Generation, 399 tok/s Prefill, TTFT 41.010 ms (3 Laufe)ASUSTeK COMPUTER INC....Dell Inc. PowerEdge R820 - 0,8 tok/s Generation, 22 tok/s Prefill, TTFT 98.716 ms (1 Lauf)Dell Inc. PowerEdge R...ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI - 6,0 tok/s Generation, 65 tok/s Prefill, TTFT 35.247 ms (1 Lauf) | DIESER LAUF★ ASUSTeK COMPUTER INC....
ASUSTeK COMPUTER INC. GX10 103,5 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 7,0 tok/s★ ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 6,0 tok/s this runDell Inc. PowerEdge R820 0,8 tok/s

ENGby engine

133100,066,733,30,001.0902.1793.269Prefill (tok/s)Generation (tok/s)vLLM - 103,5 tok/s Generation, 2.643 tok/s Prefill, TTFT 22.003 ms (6 Laufe)vLLMllama.cpp - 6,0 tok/s Generation, 43 tok/s Prefill, TTFT 66.981 ms (2 Laufe) | DIESER LAUF★ llama.cpp
vLLM 103,5 tok/s★ llama.cpp 6,0 tok/s this run

DRVby driver

13399,866,633,30,002.0044.0096.013Prefill (tok/s)Generation (tok/s)NVIDIA 590.48.01 / CUDA 13.1 - 103,5 tok/s Generation, 4.887 tok/s Prefill, TTFT 2.995 ms (3 Laufe)NVIDIA 590.48.01 / CU...unbekannt - 7,0 tok/s Generation, 257 tok/s Prefill, TTFT 51.399 ms (5 Laufe)unbekannt
NVIDIA 590.48.01 / CUDA 13.1 103,5 tok/sunbekannt 7,0 tok/s
💰 Economics

Economics of this run

Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (1× 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 70 W
⚡ TDP 644 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)644 W estimated (TDP)GPU 600 + CPU 29 + Board 15 W full load
Avg cost / hourEUR 0.19
Electricity / 1M tokensEUR 8.91
Token / kWh33.67K
Acquisition (system)EUR 15,248 full priceGPU EUR 13,000 · CPU EUR 649 · Board EUR 499 · RAM EUR 920 · PSU EUR 180
Electricity (2 years)
TCO (2 years)EUR 18,632
Output tokens (2 years)379.69M
☁️ 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 (70 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

MiniMax-M2.5NVIDIA RTX PRO 6000 Blackwell Workstation EditionMiniMax-M2.5NVIDIA GB10 (DGX Spark)MiniMax-M2.5AMD Radeon AI PRO R9700MiniMax-M2.5Keine GPU (CPU-only)
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.