Contributed byMario AlkaQwen (Alibaba) Qwen3.8-Flash-Next
Harness benchmark · measured on 08.10.2026 07:59
Benchmark-ID
run-20261023-180239-2b7b40Kolibri-Qwen Diverse Deterministic 310 V2.0MoE180BRuntime: deterministic310Quantisierung: NVFP4
ℹ
Public note: Variante: Uncensored (mazinb).
⚠
Configuration note: Es sind 2 Grafikkarten installiert, aber nur 1 wurde fuer diesen Lauf genutzt. Mit allen 2 GPUs koennte der Durchsatz deutlich hoeher liegen.
Points297,00von 310,00
Points rate95,81%
Tasks passed297von 310
Total duration2.578,63s
Ranking in the field
11of 56 systems
Harness benchmark · Primary metric: Score rate (%)
This run is better than 82 % of all comparable systems.
Score rate
96 %
+48 % vs Ø 64,9
Tasks passed
297
+243 % vs Ø 86,5
Total duration
2.578,6 s
+65 % vs Ø 1.562,5
Distribution in the field10 – 100 %
Wie schlägt sich dieser Benchmark mit anderen Modellen?
Qwen3.8-Flash-Next same model2× 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261023-180243-30647c
100 %
Kolibri-1NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261008-165122-c5dea7
100 %
Qwen3.6-35B-A3B-AWQNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260929-153935-691df9
100 %
Jev-1.13.0Cloud/API: TypeSafe AI (Hardware unbekannt) · run-20260929-091917-060aae
100 %
Qwen3.6-35B-A3B-AWQNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260929-091748-cbe456
100 %
Qwen3.8-27B-Uncensored2× 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260929-091747-e13da0
100 %
Open-Jev-27B-v1.1NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260929-091747-5d8477
100 %
Kolibri-1NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261008-125923-3f1672
99 %
Qwen3.8-Flash-Next same model2× 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261023-180239-3519b5
98 %
Qwen3.6-35B-A3B-AWQNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260929-112249-dba0f2
97 %
Qwen3.8-Flash-Next this run2× 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261023-180239-2b7b40
96 %
Kolibri-1NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261008-125923-43651d
93 %
Kolibri-1NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20261008-125923-5b5304
91 %
Gemma-4-26B-A4B-itNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260722-161611-05d866
91 %
Hardware
GPU: 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · 96 GB VRAM
CPU: AMD Ryzen Threadripper PRO 9965WX 24-Cores
RAM: 125 GB
Mainboard: ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE
Setup
Runtime: deterministic310
Quantization: NVFP4
Model: Qwen3.8-Flash-Next
Aufgaben (310)
| Task | Status | Points | Duration |
|---|---|---|---|
| arithmetic-01 | passed | 1 / 1 | 1,46 s |
| arithmetic-02 | passed | 1 / 1 | 2,63 s |
| arithmetic-03 | passed | 1 / 1 | 3,93 s |
| arithmetic-04 | passed | 1 / 1 | 1,99 s |
| arithmetic-05 | passed | 1 / 1 | 2,46 s |
| arithmetic-06 | passed | 1 / 1 | 1,72 s |
| arithmetic-07 | passed | 1 / 1 | 2,67 s |
| arithmetic-08 | passed | 1 / 1 | 2,45 s |
| arithmetic-09 | passed | 1 / 1 | 1,94 s |
| arithmetic-10 | passed | 1 / 1 | 1,58 s |
| arithmetic-11 | passed | 1 / 1 | 1,63 s |
| arithmetic-12 | passed | 1 / 1 | 2,98 s |
| arithmetic-13 | passed | 1 / 1 | 2,98 s |
| arithmetic-14 | passed | 1 / 1 | 2,18 s |
| arithmetic-15 | passed | 1 / 1 | 1,66 s |
| ordering_logic-01 | passed | 1 / 1 | 6,34 s |
| ordering_logic-02 | passed | 1 / 1 | 5,49 s |
| ordering_logic-03 | passed | 1 / 1 | 5,21 s |
| ordering_logic-04 | passed | 1 / 1 | 5,50 s |
| ordering_logic-05 | passed | 1 / 1 | 6,50 s |
| ordering_logic-06 | passed | 1 / 1 | 6,11 s |
| ordering_logic-07 | passed | 1 / 1 | 5,87 s |
| ordering_logic-08 | passed | 1 / 1 | 6,64 s |
| ordering_logic-09 | passed | 1 / 1 | 6,50 s |
| ordering_logic-10 | passed | 1 / 1 | 5,86 s |
| ordering_logic-11 | passed | 1 / 1 | 6,32 s |
| ordering_logic-12 | passed | 1 / 1 | 6,12 s |
| ordering_logic-13 | passed | 1 / 1 | 6,14 s |
| ordering_logic-14 | passed | 1 / 1 | 6,31 s |
| ordering_logic-15 | passed | 1 / 1 | 10,45 s |
| extraction-01 | passed | 1 / 1 | 2,61 s |
| extraction-02 | passed | 1 / 1 | 3,31 s |
| extraction-03 | passed | 1 / 1 | 3,46 s |
| extraction-04 | passed | 1 / 1 | 2,53 s |
| extraction-05 | passed | 1 / 1 | 3,37 s |
| extraction-06 | passed | 1 / 1 | 2,58 s |
| extraction-07 | passed | 1 / 1 | 2,36 s |
| extraction-08 | passed | 1 / 1 | 2,74 s |
| extraction-09 | passed | 1 / 1 | 2,54 s |
| extraction-10 | passed | 1 / 1 | 2,54 s |
| extraction-11 | passed | 1 / 1 | 4,37 s |
| extraction-12 | passed | 1 / 1 | 2,41 s |
| extraction-13 | passed | 1 / 1 | 2,39 s |
| extraction-14 | passed | 1 / 1 | 2,83 s |
| extraction-15 | passed | 1 / 1 | 3,20 s |
| code_comprehension-01 | passed | 1 / 1 | 1,89 s |
| code_comprehension-02 | passed | 1 / 1 | 3,34 s |
| code_comprehension-03 | passed | 1 / 1 | 9,62 s |
| code_comprehension-04 | passed | 1 / 1 | 4,26 s |
| code_comprehension-05 | passed | 1 / 1 | 5,07 s |
| code_comprehension-06 | passed | 1 / 1 | 1,65 s |
| code_comprehension-07 | passed | 1 / 1 | 3,82 s |
| code_comprehension-08 | passed | 1 / 1 | 11,57 s |
| code_comprehension-09 | passed | 1 / 1 | 5,68 s |
| code_comprehension-10 | passed | 1 / 1 | 6,25 s |
| code_comprehension-11 | passed | 1 / 1 | 1,57 s |
| code_comprehension-12 | passed | 1 / 1 | 4,03 s |
| code_comprehension-13 | passed | 1 / 1 | 7,04 s |
| code_comprehension-14 | passed | 1 / 1 | 6,22 s |
| code_comprehension-15 | passed | 1 / 1 | 7,10 s |
| grounded_qa-01 | passed | 1 / 1 | 2,27 s |
| grounded_qa-02 | passed | 1 / 1 | 1,52 s |
| grounded_qa-03 | passed | 1 / 1 | 1,37 s |
| grounded_qa-04 | passed | 1 / 1 | 2,49 s |
| grounded_qa-05 | passed | 1 / 1 | 2,73 s |
| grounded_qa-06 | passed | 1 / 1 | 2,68 s |
| grounded_qa-07 | passed | 1 / 1 | 2,11 s |
| grounded_qa-08 | passed | 1 / 1 | 1,76 s |
| grounded_qa-09 | passed | 1 / 1 | 1,30 s |
| grounded_qa-10 | passed | 1 / 1 | 1,88 s |
| grounded_qa-11 | passed | 1 / 1 | 2,34 s |
| grounded_qa-12 | passed | 1 / 1 | 1,44 s |
| grounded_qa-13 | passed | 1 / 1 | 2,10 s |
| grounded_qa-14 | passed | 1 / 1 | 1,81 s |
| grounded_qa-15 | passed | 1 / 1 | 1,49 s |
| classification-01 | failed | 0 / 1 | 4,22 s |
| classification-02 | passed | 1 / 1 | 4,77 s |
| classification-03 | passed | 1 / 1 | 4,00 s |
| classification-04 | passed | 1 / 1 | 4,80 s |
| classification-05 | passed | 1 / 1 | 3,95 s |
| classification-06 | passed | 1 / 1 | 4,08 s |
| classification-07 | passed | 1 / 1 | 3,91 s |
| classification-08 | passed | 1 / 1 | 4,24 s |
| classification-09 | passed | 1 / 1 | 4,02 s |
| classification-10 | passed | 1 / 1 | 4,02 s |
| classification-11 | passed | 1 / 1 | 4,29 s |
| classification-12 | passed | 1 / 1 | 3,90 s |
| classification-13 | passed | 1 / 1 | 4,18 s |
| classification-14 | passed | 1 / 1 | 4,99 s |
| classification-15 | passed | 1 / 1 | 4,15 s |
| path_planning-01 | passed | 1 / 1 | 4,88 s |
| path_planning-02 | passed | 1 / 1 | 5,23 s |
| path_planning-03 | failed | 0 / 1 | 5,02 s |
| path_planning-04 | passed | 1 / 1 | 5,71 s |
| path_planning-05 | passed | 1 / 1 | 5,28 s |
| path_planning-06 | passed | 1 / 1 | 5,29 s |
| path_planning-07 | passed | 1 / 1 | 5,20 s |
| path_planning-08 | passed | 1 / 1 | 5,82 s |
| path_planning-09 | passed | 1 / 1 | 5,11 s |
| path_planning-10 | passed | 1 / 1 | 5,26 s |
| path_planning-11 | passed | 1 / 1 | 4,96 s |
| path_planning-12 | passed | 1 / 1 | 5,09 s |
| path_planning-13 | passed | 1 / 1 | 5,08 s |
| path_planning-14 | passed | 1 / 1 | 5,10 s |
| path_planning-15 | passed | 1 / 1 | 4,98 s |
| relational_query-01 | passed | 1 / 1 | 5,27 s |
| relational_query-02 | passed | 1 / 1 | 4,44 s |
| relational_query-03 | passed | 1 / 1 | 7,16 s |
| relational_query-04 | passed | 1 / 1 | 4,39 s |
| relational_query-05 | passed | 1 / 1 | 4,55 s |
| relational_query-06 | passed | 1 / 1 | 4,76 s |
| relational_query-07 | passed | 1 / 1 | 5,21 s |
| relational_query-08 | passed | 1 / 1 | 4,52 s |
| relational_query-09 | passed | 1 / 1 | 4,38 s |
| relational_query-10 | passed | 1 / 1 | 3,75 s |
| relational_query-11 | passed | 1 / 1 | 4,56 s |
| relational_query-12 | passed | 1 / 1 | 4,40 s |
| relational_query-13 | passed | 1 / 1 | 4,52 s |
| relational_query-14 | passed | 1 / 1 | 4,33 s |
| relational_query-15 | passed | 1 / 1 | 4,71 s |
| date_time-01 | passed | 1 / 1 | 4,29 s |
| date_time-02 | passed | 1 / 1 | 4,11 s |
| date_time-03 | failed | 0 / 1 | 41,52 s |
| date_time-04 | passed | 1 / 1 | 5,59 s |
| date_time-05 | passed | 1 / 1 | 3,12 s |
| date_time-06 | passed | 1 / 1 | 42,18 s |
| date_time-07 | passed | 1 / 1 | 6,92 s |
| date_time-08 | passed | 1 / 1 | 3,48 s |
| date_time-09 | passed | 1 / 1 | 32,99 s |
| date_time-10 | passed | 1 / 1 | 7,66 s |
| date_time-11 | passed | 1 / 1 | 2,79 s |
| date_time-12 | failed | 0 / 1 | 45,21 s |
| date_time-13 | passed | 1 / 1 | 8,39 s |
| date_time-14 | passed | 1 / 1 | 3,78 s |
| date_time-15 | failed | 0 / 1 | 18,84 s |
| units-01 | passed | 1 / 1 | 2,29 s |
| units-02 | passed | 1 / 1 | 2,31 s |
| units-03 | passed | 1 / 1 | 1,78 s |
| units-04 | passed | 1 / 1 | 2,32 s |
| units-05 | passed | 1 / 1 | 1,66 s |
| units-06 | passed | 1 / 1 | 1,71 s |
| units-07 | passed | 1 / 1 | 2,53 s |
| units-08 | passed | 1 / 1 | 2,22 s |
| units-09 | passed | 1 / 1 | 2,55 s |
| units-10 | passed | 1 / 1 | 2,69 s |
| units-11 | passed | 1 / 1 | 2,34 s |
| units-12 | passed | 1 / 1 | 2,61 s |
| units-13 | passed | 1 / 1 | 1,76 s |
| units-14 | passed | 1 / 1 | 2,28 s |
| units-15 | passed | 1 / 1 | 2,03 s |
| string_processing-01 | passed | 1 / 1 | 2,28 s |
| string_processing-02 | passed | 1 / 1 | 2,40 s |
| string_processing-03 | passed | 1 / 1 | 2,40 s |
| string_processing-04 | passed | 1 / 1 | 2,23 s |
| string_processing-05 | passed | 1 / 1 | 2,29 s |
| string_processing-06 | passed | 1 / 1 | 2,47 s |
| string_processing-07 | passed | 1 / 1 | 2,26 s |
| string_processing-08 | passed | 1 / 1 | 2,44 s |
| string_processing-09 | passed | 1 / 1 | 2,35 s |
| string_processing-10 | passed | 1 / 1 | 2,03 s |
| string_processing-11 | passed | 1 / 1 | 2,37 s |
| string_processing-12 | passed | 1 / 1 | 2,60 s |
| string_processing-13 | passed | 1 / 1 | 2,32 s |
| string_processing-14 | passed | 1 / 1 | 2,36 s |
| string_processing-15 | passed | 1 / 1 | 2,36 s |
| sets-01 | passed | 1 / 1 | 3,05 s |
| sets-02 | passed | 1 / 1 | 4,54 s |
| sets-03 | passed | 1 / 1 | 8,76 s |
| sets-04 | passed | 1 / 1 | 4,01 s |
| sets-05 | passed | 1 / 1 | 5,71 s |
| sets-06 | passed | 1 / 1 | 5,60 s |
| sets-07 | passed | 1 / 1 | 3,90 s |
| sets-08 | passed | 1 / 1 | 5,29 s |
| sets-09 | passed | 1 / 1 | 4,88 s |
| sets-10 | passed | 1 / 1 | 3,40 s |
| sets-11 | passed | 1 / 1 | 4,89 s |
| sets-12 | passed | 1 / 1 | 8,96 s |
| sets-13 | passed | 1 / 1 | 2,85 s |
| sets-14 | passed | 1 / 1 | 5,14 s |
| sets-15 | passed | 1 / 1 | 8,17 s |
| constraint_selection-01 | passed | 1 / 1 | 4,81 s |
| constraint_selection-02 | passed | 1 / 1 | 4,47 s |
| constraint_selection-03 | passed | 1 / 1 | 4,37 s |
| constraint_selection-04 | passed | 1 / 1 | 4,23 s |
| constraint_selection-05 | passed | 1 / 1 | 4,62 s |
| constraint_selection-06 | passed | 1 / 1 | 3,84 s |
| constraint_selection-07 | passed | 1 / 1 | 4,42 s |
| constraint_selection-08 | passed | 1 / 1 | 4,91 s |
| constraint_selection-09 | passed | 1 / 1 | 4,65 s |
| constraint_selection-10 | passed | 1 / 1 | 3,87 s |
| constraint_selection-11 | passed | 1 / 1 | 4,68 s |
| constraint_selection-12 | passed | 1 / 1 | 3,89 s |
| constraint_selection-13 | passed | 1 / 1 | 3,74 s |
| constraint_selection-14 | passed | 1 / 1 | 4,18 s |
| constraint_selection-15 | passed | 1 / 1 | 4,19 s |
| probability-01 | passed | 1 / 1 | 3,44 s |
| probability-02 | passed | 1 / 1 | 2,54 s |
| probability-03 | passed | 1 / 1 | 2,03 s |
| probability-04 | passed | 1 / 1 | 2,49 s |
| probability-05 | passed | 1 / 1 | 1,87 s |
| probability-06 | passed | 1 / 1 | 2,55 s |
| probability-07 | passed | 1 / 1 | 2,49 s |
| probability-08 | passed | 1 / 1 | 2,09 s |
| probability-09 | passed | 1 / 1 | 2,50 s |
| probability-10 | passed | 1 / 1 | 2,09 s |
| probability-11 | passed | 1 / 1 | 2,10 s |
| probability-12 | passed | 1 / 1 | 2,30 s |
| probability-13 | passed | 1 / 1 | 1,54 s |
| probability-14 | passed | 1 / 1 | 2,08 s |
| probability-15 | passed | 1 / 1 | 2,28 s |
| ext-bit_circuits-01 | passed | 1 / 1 | 7,55 s |
| ext-bit_circuits-02 | passed | 1 / 1 | 6,46 s |
| ext-bit_circuits-03 | passed | 1 / 1 | 13,14 s |
| ext-bit_circuits-04 | passed | 1 / 1 | 5,28 s |
| ext-bit_circuits-05 | passed | 1 / 1 | 7,56 s |
| ext-bit_circuits-06 | passed | 1 / 1 | 6,44 s |
| ext-bit_circuits-07 | passed | 1 / 1 | 8,43 s |
| ext-bit_circuits-08 | passed | 1 / 1 | 5,89 s |
| ext-bit_circuits-09 | passed | 1 / 1 | 7,27 s |
| ext-bit_circuits-10 | passed | 1 / 1 | 6,27 s |
| ext-weighted_scheduling-01 | passed | 1 / 1 | 22,11 s |
| ext-weighted_scheduling-02 | passed | 1 / 1 | 22,46 s |
| ext-weighted_scheduling-03 | failed | 0 / 1 | 66,09 s |
| ext-weighted_scheduling-04 | passed | 1 / 1 | 15,69 s |
| ext-weighted_scheduling-05 | passed | 1 / 1 | 32,74 s |
| ext-weighted_scheduling-06 | passed | 1 / 1 | 16,62 s |
| ext-weighted_scheduling-07 | passed | 1 / 1 | 30,35 s |
| ext-weighted_scheduling-08 | passed | 1 / 1 | 17,95 s |
| ext-weighted_scheduling-09 | passed | 1 / 1 | 7,94 s |
| ext-weighted_scheduling-10 | passed | 1 / 1 | 19,86 s |
| ext-finite_automata-01 | passed | 1 / 1 | 4,50 s |
| ext-finite_automata-02 | passed | 1 / 1 | 6,01 s |
| ext-finite_automata-03 | passed | 1 / 1 | 8,00 s |
| ext-finite_automata-04 | passed | 1 / 1 | 6,06 s |
| ext-finite_automata-05 | passed | 1 / 1 | 4,93 s |
| ext-finite_automata-06 | passed | 1 / 1 | 3,26 s |
| ext-finite_automata-07 | passed | 1 / 1 | 5,53 s |
| ext-finite_automata-08 | failed | 0 / 1 | 29,65 s |
| ext-finite_automata-09 | passed | 1 / 1 | 4,65 s |
| ext-finite_automata-10 | passed | 1 / 1 | 5,13 s |
| ext-fifo_inventory-01 | passed | 1 / 1 | 4,77 s |
| ext-fifo_inventory-02 | passed | 1 / 1 | 4,85 s |
| ext-fifo_inventory-03 | passed | 1 / 1 | 5,18 s |
| ext-fifo_inventory-04 | passed | 1 / 1 | 5,29 s |
| ext-fifo_inventory-05 | passed | 1 / 1 | 5,09 s |
| ext-fifo_inventory-06 | passed | 1 / 1 | 4,74 s |
| ext-fifo_inventory-07 | passed | 1 / 1 | 4,80 s |
| ext-fifo_inventory-08 | passed | 1 / 1 | 5,34 s |
| ext-fifo_inventory-09 | passed | 1 / 1 | 5,49 s |
| ext-fifo_inventory-10 | passed | 1 / 1 | 5,39 s |
| ext-edit_distance-01 | passed | 1 / 1 | 47,28 s |
| ext-edit_distance-02 | failed | 0 / 1 | 64,97 s |
| ext-edit_distance-03 | passed | 1 / 1 | 44,36 s |
| ext-edit_distance-04 | failed | 0 / 1 | 64,82 s |
| ext-edit_distance-05 | passed | 1 / 1 | 59,79 s |
| ext-edit_distance-06 | failed | 0 / 1 | 65,02 s |
| ext-edit_distance-07 | passed | 1 / 1 | 47,63 s |
| ext-edit_distance-08 | passed | 1 / 1 | 62,59 s |
| ext-edit_distance-09 | passed | 1 / 1 | 52,94 s |
| ext-edit_distance-10 | passed | 1 / 1 | 59,11 s |
| ext-calendar_explicit-01 | passed | 1 / 1 | 6,68 s |
| ext-calendar_explicit-02 | passed | 1 / 1 | 8,59 s |
| ext-calendar_explicit-03 | passed | 1 / 1 | 13,48 s |
| ext-calendar_explicit-04 | passed | 1 / 1 | 23,81 s |
| ext-calendar_explicit-05 | passed | 1 / 1 | 8,85 s |
| ext-calendar_explicit-06 | passed | 1 / 1 | 9,22 s |
| ext-calendar_explicit-07 | passed | 1 / 1 | 19,21 s |
| ext-calendar_explicit-08 | passed | 1 / 1 | 27,09 s |
| ext-calendar_explicit-09 | passed | 1 / 1 | 20,13 s |
| ext-calendar_explicit-10 | passed | 1 / 1 | 10,41 s |
| ext-rolling_checksum-01 | passed | 1 / 1 | 12,44 s |
| ext-rolling_checksum-02 | passed | 1 / 1 | 15,27 s |
| ext-rolling_checksum-03 | passed | 1 / 1 | 8,30 s |
| ext-rolling_checksum-04 | passed | 1 / 1 | 12,99 s |
| ext-rolling_checksum-05 | passed | 1 / 1 | 7,12 s |
| ext-rolling_checksum-06 | passed | 1 / 1 | 13,73 s |
| ext-rolling_checksum-07 | passed | 1 / 1 | 11,57 s |
| ext-rolling_checksum-08 | passed | 1 / 1 | 15,16 s |
| ext-rolling_checksum-09 | passed | 1 / 1 | 13,85 s |
| ext-rolling_checksum-10 | passed | 1 / 1 | 14,01 s |
| ext-grid_path_count-01 | passed | 1 / 1 | 8,99 s |
| ext-grid_path_count-02 | passed | 1 / 1 | 7,10 s |
| ext-grid_path_count-03 | passed | 1 / 1 | 7,72 s |
| ext-grid_path_count-04 | passed | 1 / 1 | 9,18 s |
| ext-grid_path_count-05 | passed | 1 / 1 | 11,70 s |
| ext-grid_path_count-06 | passed | 1 / 1 | 8,35 s |
| ext-grid_path_count-07 | passed | 1 / 1 | 6,64 s |
| ext-grid_path_count-08 | passed | 1 / 1 | 10,20 s |
| ext-grid_path_count-09 | passed | 1 / 1 | 7,20 s |
| ext-grid_path_count-10 | passed | 1 / 1 | 7,11 s |
| ext-sample_variance-01 | passed | 1 / 1 | 10,18 s |
| ext-sample_variance-02 | passed | 1 / 1 | 5,84 s |
| ext-sample_variance-03 | passed | 1 / 1 | 16,74 s |
| ext-sample_variance-04 | passed | 1 / 1 | 12,67 s |
| ext-sample_variance-05 | passed | 1 / 1 | 13,34 s |
| ext-sample_variance-06 | passed | 1 / 1 | 13,28 s |
| ext-sample_variance-07 | passed | 1 / 1 | 5,40 s |
| ext-sample_variance-08 | passed | 1 / 1 | 15,98 s |
| ext-sample_variance-09 | passed | 1 / 1 | 18,65 s |
| ext-sample_variance-10 | passed | 1 / 1 | 8,04 s |
| ext-round_robin-01 | failed | 0 / 1 | 13,43 s |
| ext-round_robin-02 | passed | 1 / 1 | 7,21 s |
| ext-round_robin-03 | passed | 1 / 1 | 8,03 s |
| ext-round_robin-04 | passed | 1 / 1 | 11,29 s |
| ext-round_robin-05 | failed | 0 / 1 | 7,35 s |
| ext-round_robin-06 | passed | 1 / 1 | 7,46 s |
| ext-round_robin-07 | passed | 1 / 1 | 19,58 s |
| ext-round_robin-08 | failed | 0 / 1 | 7,08 s |
| ext-round_robin-09 | passed | 1 / 1 | 12,25 s |
| ext-round_robin-10 | passed | 1 / 1 | 14,39 s |
Configuration
# LLM-Benchmark Konfiguration
# Modell : Qwen3.8-Flash-Next
# Engine : vLLM
# Run-ID : run-20261023-180239-2b7b40
# GPU : 2x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
# CPU : AMD Ryzen Threadripper PRO 9965WX 24-Cores
# RAM : 125 GB
bench@llm-benchmark:~$ /usr/bin/python3 /usr/local/bin/vllm serve mazinb/Qwen3.8-Flash-Next-Uncensored-NVFP4 \
--served-model-name Qwen3.8-Flash-Next-Uncensored \
--host 0.0.0.0 \
--port 8000 \
--trust-remote-code \
--distributed-executor-backend mp \
--gpu-memory-utilization 0.90 \
--max-model-len 131072 \
--max-num-seqs 10 \
--enable-auto-tool-choice \
--tool-call-parser qwen3_coder \
--reasoning-parser qwen3
| 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. | vllm |
| Modellalias?Der Name, unter dem das Modell ueber die API angesprochen wird. Genau dieser Wert muss im Request-Feld 'model' stehen. | Qwen3.8-Flash-Next-Uncensored |
| Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann. | 131072 |
| Alias?Anzeigename des Modells nach aussen (served model name), unabhaengig vom Dateinamen. | Qwen3.8-Flash-Next-Uncensored |
| Trust-Remote-Code?Erlaubt vLLM, mit dem Modell mitgelieferten Python-Code auszufuehren. Noetig fuer Architekturen, die vLLM nicht von Haus aus kennt. Nur bei vertrauenswuerdigen Modellen aktivieren. | aktiv |
| Executor-Backend?Verfahren fuer Multi-GPU/Multi-Node: mp = Prozesse auf einer Maschine, ray = ueber Ray (auch ueber mehrere Knoten). | mp |
| GPU-Speicher?Anteil des GPU-Speichers (0 bis 1), den vLLM belegen darf. 0.92 = 92 %. Hoeher = mehr Platz fuer den KV-Cache (mehr/laengere parallele Anfragen), aber groesseres Risiko fuer 'Out of Memory'. | 0.90 |
| Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells. | 131072 |
| Max-Sequenzen?Maximale Anzahl gleichzeitig bearbeiteter Anfragen (Batch-Groesse). Hoeher = mehr Durchsatz, braucht aber mehr KV-Cache-Speicher. | 10 |
| Auto-Tool-Choice?Erlaubt dem Modell, Tools (Function-Calling) selbststaendig auszuwaehlen. MUSS aktiv sein, damit 'tools' im Request akzeptiert werden - sonst lehnt vLLM sie mit HTTP 400 ab. | aktiv |
| Tool-Parser?Bestimmt, wie vLLM Tool-Aufrufe aus der Modellantwort ausliest (z.B. hermes, mistral, llama3_json). Muss zum Modell passen. | qwen3_coder |
| Reasoning-Parser?Trennt den Denk-/Reasoning-Teil der Antwort vom eigentlichen Inhalt (z.B. bei Modellen mit <think>-Bloecken). | qwen3 |
| 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
★ NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.564,6 tok/s this runNVIDIA RTX A6000 1.171,8 tok/sNVIDIA GeForce RTX 3090 1.132,7 tok/sNVIDIA RTX PRO 6000 Blackwell Workstation Edition 584,1 tok/sNVIDIA GB10 (DGX Spark) 492,8 tok/sNVIDIA Tesla V100 PCIe 32GB 196,2 tok/sAMD Radeon RX 7900 XTX 18,7 tok/s
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/sAMD Ryzen 9 9950X 16-Core Processor 584,1 tok/sNVIDIA Grace 492,8 tok/s★ AMD Ryzen Threadripper PRO 9965WX 24-Cores 424,2 tok/s this runAMD Ryzen 3 PRO 3200GE w/ Radeon Vega Graphics 196,2 tok/sAMD EPYC 7203P 8-Core Processor 79,2 tok/sIntel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz 63,2 tok/sAMD 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. ProArt X870E-CREATOR WIFI 584,1 tok/sASUSTeK COMPUTER INC. GX10 492,8 tok/s★ ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 424,2 tok/s this runASUSTeK COMPUTER INC. ROG STRIX X570-F GAMING 196,2 tok/sASRockRack ROMED8-2T/BCM 79,2 tok/sDell Inc. PowerEdge R820 63,2 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 35,4 tok/s
ENGby engine
DRVby driver
unbekannt 1.564,6 tok/sNVIDIA 590.48.01 / CUDA 13.1 492,8 tok/s
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.