BridgeBenchBridgeBench

Qwen 3.5 27B

Rank #1 · 27B · FP16

Summary

Pass Rate

73.1%

Tasks Passed

76/104

Model Size

27B

Quantization

FP16

Median Throughput

11.1 tok/s

Median TTFT

361 ms

Inference Success

86.7%

Avg Latency

182225 ms

Hardware Profile

Device

DGX Spark

Chip

GB10 Grace Blackwell

Memory

128 GB Unified

Backend

ollama

Quantization

FP16

Peak GPU Mem

0.0 GB

Category Results

Speed
11.1 tok/s · 361ms TTFT
Hallucination
12/30
40.0%
Code Generation
15/20
75.0%
Reasoning
19/20
95.0%
Instruction Following
10/14
71.4%

Task Results

Speed20/20 passed
TaskDifficultyResultLatencyTokens
ttft-short-100standardPass9165ms100
ttft-short-200standardPass9199ms100
ttft-medium-500standardPass9212ms100
ttft-medium-1kstandardPass9329ms100
ttft-long-2kstandardPass9963ms100
ttft-chat-contextstandardPass9225ms100
ttft-json-outputstandardPass9242ms100
ttft-multilangstandardPass9321ms100
ttft-reasoningstandardPass9279ms100
ttft-creativestandardPass9245ms100
tp-essaystandardPass180398ms2000
tp-code-appstandardPass180322ms2000
tp-tutorialstandardPass135105ms1500
tp-analysisstandardPass135165ms1500
tp-debugstandardPass135474ms1500
tp-architecturestandardPass180302ms2000
tp-comparisonstandardPass180300ms2000
tp-securitystandardPass180359ms2000
tp-algorithmstandardPass180421ms2000
tp-documentationstandardPass180376ms2000
Hallucination12/30 passed
TaskDifficultyResultLatencyTokens
fact-01easyPass47453ms526
fact-02mediumPass37150ms411
fact-03mediumFail42435ms470
fact-04hardPass40909ms453
fact-05hardFail87840ms975
fact-06easyPass49541ms549
fact-07mediumPass35822ms396
fact-08hardPass37728ms417
fact-09mediumPass22969ms253
fact-10hardPass105456ms1171
code-01easyFail35094ms388
code-02mediumFail37409ms414
code-03mediumFail104976ms1165
code-04hardFail50737ms561
code-05hardFail70253ms779
code-06easyFail60911ms675
code-07mediumPass64199ms712
code-08hardPass91541ms1016
code-09mediumFail83442ms926
code-10hardPass53601ms594
cal-01mediumFail42064ms466
cal-02hardFail112964ms1254
cal-03mediumFail131965ms1465
cal-04easyFail65535ms726
cal-05hardFail190942ms2119
cal-06mediumFail63206ms701
cal-07hardFail348529ms3858
cal-08mediumFail42718ms473
cal-09easyPass31136ms344
cal-10hardFail47216ms523
Code Generation15/20 passed
TaskDifficultyResultLatencyTokens
fn-01easyPass84327ms935
fn-02easyPass313189ms3468
fn-03mediumFail87215ms967
fn-04mediumPass354018ms3917
fn-05mediumPass254742ms2822
fn-06hardFail370301ms4096
fn-07hardPass121735ms1350
fn-08hardFail370428ms4096
bug-01easyPass69583ms770
bug-02mediumPass245621ms2719
bug-03hardPass370652ms4096
bug-04mediumPass134955ms1495
algo-01mediumPass319326ms3536
algo-02hardPass237490ms2632
algo-03mediumPass157073ms1742
algo-04hardPass129459ms1436
multi-01hardFail370323ms4096
multi-02hardFail370276ms4096
multi-03hardPass204084ms2263
multi-04hardPass333035ms3686
Reasoning19/20 passed
TaskDifficultyResultLatencyTokens
arith-01hardPass131219ms668
arith-02hardPass262922ms460
arith-03expertPass274552ms550
arith-04expertPass192608ms1037
arith-05expertPass217093ms289
arith-06hardError300891ms0
spatial-01hardPass242588ms505
spatial-02expertError300911ms0
spatial-03expertError300908ms0
spatial-04hardPass197701ms655
spatial-05expertError300936ms0
spatial-06hardPass485585ms4096
cstr-01hardPass333972ms682
cstr-02expertError300914ms0
cstr-03expertPass597644ms4096
cstr-04hardError300907ms0
cstr-05expertError300902ms0
cstr-06hardError600001ms0
adv-01hardError300878ms0
adv-02expertPass546274ms4096
adv-03expertPass213678ms332
adv-04hardError300936ms0
adv-05expertPass96946ms306
adv-06expertPass161007ms841
cf-01hardPass74787ms371
cf-02expertFail118635ms607
cf-03expertPass205478ms1098
cf-04hardPass110954ms965
cf-05expertPass210882ms970
cf-06expertPass120444ms563
Instruction Following10/14 passed
TaskDifficultyResultLatencyTokens
fmt-01easyPass223005ms1090
fmt-02easyPass240725ms844
fmt-03mediumPass232298ms374
fmt-04mediumPass227379ms761
fmt-05hardFail518861ms4096
fmt-06hardFail543421ms4096
con-01easyError300905ms0
con-02easyPass173987ms506
con-03mediumError300921ms0
con-04mediumError300916ms0
con-05hardPass411547ms3019
con-06hardPass589605ms4096
role-01mediumError300901ms0
role-02mediumPass392726ms1090
role-03hardPass543323ms2811
role-04hardError300896ms0
mc-01hardFail511208ms4096
mc-02hardError300916ms0
mc-03hardPass253132ms2037
mc-04hardFail400276ms4096