Contributed byMarcel SommerQwen (Alibaba) Qwen3.8-Flash-Next
Performance benchmark · measured on 21.10.2026 09:48
Benchmark-ID
run-20261021-093816-154b4eTimebench 3 - Kombi (Prefill + Generation)MoE180BRuntime: strataQuantisierung: IQ3_S
Generation59,28tok/s
Prefill483,09tok/s
Time to First Token4.083,50ms
Total duration42,72s
Concurrency1parallel
Ranking in the field
8of 14 systems
Performance benchmark · Primary metric: Generation-Speed (tok/s) · 1× concurrent
This run is better than 46 % of all comparable systems.
Generation
59,3 tok/s
-16 % vs Ø 70,2
Prefill
483,1 tok/s
-74 % vs Ø 1.828,0
Time to First Token
4.084 ms
+58 % vs Ø 2.591
Distribution in the field2 – 132 tok/s
Wie schlägt sich dieser Benchmark mit anderen Modellen?
gpt-oss-20bNVIDIA Tesla V100 PCIe 32GB · run-20260907-133945-d59404
132,3 tok/s
gemma-4-E2B-itNVIDIA Tesla V100 PCIe 32GB · run-20260907-133945-5f69da
129,0 tok/s
gemma-4-E2B-itNVIDIA Tesla V100 PCIe 32GB · run-20260907-123328-fcdf51
118,0 tok/s
Qwen3.6-35B-A3BNVIDIA Tesla V100 PCIe 32GB · run-20260907-155130-2f9e6f
94,4 tok/s
Ornith-1.0-9BNVIDIA Tesla V100 PCIe 32GB · run-20260907-144537-969e36
83,8 tok/s
Meta-Llama-3.1-8B-InstructNVIDIA Tesla V100 PCIe 32GB · run-20260907-130632-9dee88
69,9 tok/s
Ornith-1.0-9BNVIDIA Tesla V100 PCIe 32GB · run-20260907-130632-76d97b
62,1 tok/s
Qwen3.8-Flash-Next this runNVIDIA Tesla V100 PCIe 32GB · run-20261021-093816-154b4e
59,3 tok/s
gemma-4-12B-itNVIDIA Tesla V100 PCIe 32GB · run-20260907-155129-1c745e
58,6 tok/s
Qwen3.8-Flash-Next same modelNVIDIA Tesla V100 PCIe 32GB · run-20261019-134138-464b84
52,1 tok/s
Qwen3.8-Flash-Next same modelNVIDIA Tesla V100 PCIe 32GB · run-20261019-100734-f18656
49,6 tok/s
gemma-4-12B-itNVIDIA Tesla V100 PCIe 32GB · run-20260907-141239-a485a6
36,4 tok/s
Ministral-3-14B-Reasoning-2512NVIDIA Tesla V100 PCIe 32GB · run-20260907-141239-8102bc
35,1 tok/s
Qwen2.5-72B-InstructNVIDIA Tesla V100 PCIe 32GB · run-20260722-165905-1dde01
2,3 tok/s
How does this benchmark compare on other GPUs?
Same model on different hardware · 1× concurrent · Generation (tok/s)
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20261012-182514-5f16dc
164,6 tok/s
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20261012-175110-537c3d
158,1 tok/s
NVIDIA RTX A6000run-20261021-093815-fdc402
125,2 tok/s
NVIDIA GeForce RTX 3090run-20261013-092430-8e3577
116,4 tok/s
NVIDIA RTX A6000run-20261021-093816-efa091
115,5 tok/s
NVIDIA Tesla V100 PCIe 32GB this runrun-20261021-093816-154b4e
59,3 tok/s
NVIDIA Tesla V100 PCIe 32GBrun-20261019-134138-464b84
52,1 tok/s
NVIDIA GeForce RTX 3090run-20261013-165111-42c54a
50,5 tok/s
NVIDIA GB10run-20261013-151008-44a477
49,9 tok/s
NVIDIA Tesla V100 PCIe 32GBrun-20261019-100734-f18656
49,6 tok/s
NVIDIA GB10run-20261013-165111-ea315f
41,8 tok/s
2× NVIDIA RTX A6000run-20261014-073140-f006db
38,8 tok/s
2× NVIDIA RTX A6000run-20260930-183837-07a63e
37,0 tok/s
NVIDIA GB10run-20261013-151009-529362
36,4 tok/s
Hardware
GPU: NVIDIA Tesla V100 PCIe 32GB · 32 GB VRAM
CPU: 4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz
RAM: 504 GB
Mainboard: Dell Inc. PowerEdge R820
Setup
Runtime: strata
Quantization: IQ3_S
Model: Qwen3.8-Flash-Next
Configuration
# LLM-Benchmark Konfiguration
# Modell : Qwen3.8-Flash-Next
# Run-ID : run-20261021-093816-154b4e
# GPU : NVIDIA Tesla V100 PCIe 32GB
# CPU : 4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz
# RAM : 504 GB
bench@llm-benchmark:~$ cat benchmark.conf
execution_type local
| execution_type | local |
Anzeige
Model comparison
Qwen3.8-Flash-Next 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
CPUby processor
AMD Ryzen 9 7945HX with Radeon Graphics 1.564,6 tok/sAMD Ryzen Threadripper PRO 3945WX 12-Cores 1.171,8 tok/sAMD Ryzen 5 5600X 6-Core Processor 1.132,7 tok/sNVIDIA Grace 492,8 tok/sAMD Ryzen 3 PRO 3200GE w/ Radeon Vega Graphics 196,2 tok/sAMD EPYC 7203P 8-Core Processor 79,2 tok/s★ Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz 63,2 tok/s this runAMD Ryzen Threadripper PRO 3955WX 16-Cores 18,7 tok/s
MBby mainboard
Shenzhen Meigao Electronic Equipment Co.,Ltd DRFXI (MotherBoard Series) 1.564,6 tok/sunbekannt 1.171,8 tok/sASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.132,7 tok/sASUSTeK COMPUTER INC. GX10 492,8 tok/sASUSTeK COMPUTER INC. ROG STRIX X570-F GAMING 196,2 tok/sASRockRack ROMED8-2T/BCM 79,2 tok/s★ Dell Inc. PowerEdge R820 63,2 tok/s this runASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 35,4 tok/s
DRVby driver
unbekannt 1.564,6 tok/sNVIDIA 590.48.01 / CUDA 13.1 492,8 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
35 W
⚡ TDP
289 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)289 W estimated (TDP)GPU 250 + CPU 29 + Board 10 W full load
Avg cost / hourEUR 0.087
Electricity / 1M tokensEUR 0.41
Token / kWh739.08K
Acquisition (system)EUR 2,795 partial priceGPU EUR 600 · CPU EUR 59 · RAM EUR 2,016 · PSU EUR 120
Electricity (2 years)–
TCO (2 years)EUR 4,313
Output tokens (2 years)3.74B
☁️ 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 (35 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
| Qwen3.8-Flash-NextNVIDIA Tesla V100 PCIe 32GB | Qwen3.8-Flash-NextNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | Qwen3.8-Flash-NextNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | Qwen3.8-Flash-NextNVIDIA RTX A6000 | |
|---|---|---|---|---|
| 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) | – | – | – | – |
NVIDIA Tesla V100 PCIe 32GB · this run 59 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 165 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 158 tok/sNVIDIA RTX A6000 125 tok/s
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.