Traces

Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.

Kolmogorov Audio Compression

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0313$38.45102.2M362k20h 1m39710
2Kimi K30.0266$26.1560.2M321k19h 29m30313
3Kimi K30.0251$50.26120.4M460k19h 45m4107
4Kimi K30.0250$34.5393.2M354k19h 44m3958
5Kimi K30.0120$62.95172.7M576k20h 0m47315

Astronomy Toolkit

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.6250$60.98152.4M478k16h 31m41412
2Kimi K30.3983$90.41238.0M505k20h 1m58720
3Kimi K30.3959$20.0846.3M279k6h 2m2055
4Kimi K30.3804$70.49167.8M452k19h 35m4688
5Kimi K30.2995$53.30137.6M418k11h 54m4307

Cranelift Codegen Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0603$148.59441.7M786k16h 22m102815
2Kimi K30.0478$126.89362.2M801k17h 53m9369
3Kimi K30.0132$145.36415.9M1.0M20h 3m10797
4Kimi K30.0000$136.80389.4M1.1M20h 0m100619
5Kimi K30.0000$136.72403.0M752k19h 8m114711

Dart Style in Haskell

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.5650$209.81607.6M1.3M20h 1m145419
2Kimi K30.5395$190.44541.2M1.2M19h 50m13363
3Kimi K30.4874$196.59547.1M1.3M19h 46m12418
4Kimi K30.4305$123.27338.1M738k19h 52m6701
5Kimi K30.0000$86.57238.4M589k20h 1m5931

FFmpeg libswscale Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.1260$127.79342.6M1.2M13h 59m82713
2Kimi K30.1252$211.22606.6M1.4M19h 43m13823
3Kimi K30.0799$127.97356.2M1.1M15h 4m91411
4Kimi K30.0639$138.03382.3M1.1M17h 32m9022
5Kimi K30.0000$171.14500.1M1.1M19h 41m117547

Crash-Proof Flash Filesystem

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0431$197.20550.4M1.5M20h 1m11971
2Kimi K30.0112$202.76569.3M1.5M19h 50m12372
3Kimi K30.0049$158.40411.0M1.7M20h 0m10101
4Kimi K30.0049$190.49508.0M1.5M19h 54m11119
5Kimi K30.0008$195.88529.5M1.7M20h 0m12662

FrogsGame Post-Training

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.4620$12.7324.9M159k19h 34m1721
2Kimi K30.3000$16.0038.1M240k20h 0m1871
3Kimi K30.1900$13.6332.0M177k20h 0m1981
4Kimi K30.1040$11.1320.9M147k19h 47m1741
5Kimi K30.0660$8.1017.2M123k19h 18m1531

Stepper Music Sequencer GBA

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.5098$230.68643.2M1.9M20h 0m15453
2Kimi K30.3706$227.60629.2M1.9M19h 47m15042
3Kimi K30.3170$227.44622.4M1.9M19h 52m14391
4Kimi K30.1336$196.38508.8M1.8M19h 51m12094
5Kimi K30.0909$184.74495.2M1.5M19h 58m11542

Git to Zig

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.1768$251.00752.4M1.2M20h 2m16191
2Kimi K30.1733$200.20602.7M965k19h 43m148313
3Kimi K30.1616$254.18761.8M1.1M19h 47m16689
4Kimi K30.1574$259.84779.9M1.1M19h 53m17652
5Kimi K30.1473$222.64660.4M1.1M19h 44m14911

Granite Mamba2 Inference Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.7872$122.78337.3M986k19h 10m77136
2Kimi K30.7805$103.11286.2M907k14h 49m65110
3Kimi K30.7571$91.87248.8M620k20h 0m53150
4Kimi K30.7477$74.48205.5M639k8h 3m45620
5Kimi K30.6169$112.73307.9M980k13h 16m70831

Higgs Uncertainty Inference

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.9593$29.5562.1M341k12h 49m28515
2Kimi K30.4200$32.1858.0M281k15h 50m2808
3Kimi K30.0960$32.3577.4M380k13h 56m31115
4Kimi K30.0626$34.0868.6M372k14h 20m27512
5Kimi K30.0289$90.08230.0M659k20h 0m5551

SGLang Inference System Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0103$110.14311.1M507k20h 1m74011
2Kimi K30.0000$185.83517.0M1.5M20h 0m118813
3Kimi K30.0000$246.45710.4M1.8M20h 0m173614
4Kimi K30.0000$49.64141.5M328k6h 19m5297
5Kimi K30.0000$124.72355.5M911k15h 41m94824

Lean 4 Kernel Type Checker in Pascal

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.9198$192.68537.9M1.4M20h 0m12181
2Kimi K30.8313$155.97405.9M1.5M19h 47m9353
3Kimi K30.5265$147.65387.7M1.4M20h 2m8941
4Kimi K30.4229$171.33477.1M1.3M19h 55m113012
5Kimi K30.2644$151.43423.8M1.1M20h 1m9281

libexpat Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.1778$238.01679.3M1.8M19h 52m158883
2Kimi K30.1513$169.56484.2M1.2M19h 34m115516
3Kimi K30.1288$173.14509.0M1.1M13h 23m114928
4Kimi K30.0000$214.44584.8M1.8M20h 0m14591
5Kimi K30.0000$256.75739.1M1.4M20h 0m15761

Lua Native Compiler

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.9575$183.51524.3M1.3M19h 59m120320
2Kimi K30.8750$185.22530.7M1.3M20h 0m12071
3Kimi K30.5063$220.93645.6M1.2M20h 2m14611
4Kimi K30.2862$159.88431.8M1.4M19h 11m10091
5Kimi K30.0000$228.55670.7M1.4M19h 8m14083

Machine-Learned Interatomic Potential

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.9282$14.8121.6M194k19h 20m1753
2Kimi K30.9177$11.7220.5M161k18h 47m1653
3Kimi K30.8916$6.549.0M95k18h 28m1262
4Kimi K30.8033$15.1121.0M176k19h 11m1683
5Kimi K30.7378$9.8813.2M119k19h 28m1432

Medium-Range Weather Forecast

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0000$20.2549.9M257k19h 31m2968
2Kimi K30.0000$16.0342.3M172k19h 35m2821
3Kimi K30.0000$13.2932.3M193k19h 22m2452
4Kimi K30.0000$11.0428.1M135k19h 31m2616
5Kimi K30.0000$19.0948.5M196k18h 52m3266

MEG Speech Decoding

task page →1 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.1102$14.9032.2M173k18h 22m2543

Wan 2.1 on MAX/Mojo

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.9307$34.4290.2M306k9h 26m3065
2Kimi K30.9196$29.2075.3M345k8h 0m2868
3Kimi K30.9090$15.0637.9M188k7h 2m2047
4Kimi K30.8899$31.5184.6M291k12h 22m3483
5Kimi K30.2838$85.29208.6M954k19h 56m5583

MS/MS De Novo Generation

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.1101$21.3546.7M229k17h 35m2826
2Kimi K30.1097$26.9062.8M258k19h 1m3257
3Kimi K30.1091$18.2139.5M209k19h 3m2395
4Kimi K30.0000$18.6638.3M209k20h 0m2544
5Kimi K30.0000$31.9877.8M334k19h 25m3394

Multi-GPU Efficient Finetuning

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.3333$19.1518.0M131k19h 9m1676
2Kimi K30.3095$16.5615.3M148k19h 10m1343
3Kimi K30.1905$16.1311.8M128k18h 47m1161
4Kimi K30.0476$25.1335.1M233k19h 37m2361
5Kimi K30.0000$10.948.7M101k18h 40m1211

Notebook Compression

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.2647$37.6696.4M363k19h 31m40917
2Kimi K30.2338$43.76111.9M405k17h 35m40312
3Kimi K30.2242$36.2787.9M412k11h 52m38212
4Kimi K30.2151$24.4560.7M274k11h 45m30210
5Kimi K30.1961$13.5921.1M194k9h 53m1799

Flight-Sim Renderer in OpenGL

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0249$126.07350.4M1.1M19h 56m9367
2Kimi K30.0214$168.60476.4M1.3M20h 1m10795
3Kimi K30.0104$167.00460.9M1.4M20h 1m10391
4Kimi K30.0079$184.18516.4M1.3M19h 37m10942
5Kimi K30.0000$182.81498.0M1.4M20h 1m10632

Optimizer Design

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.3057$15.0637.2M181k19h 11m2754
2Kimi K30.1475$41.51100.0M508k20h 1m3544
3Kimi K30.1370$16.0538.0M228k19h 25m2224
4Kimi K30.0990$22.5849.0M297k17h 53m2416
5Kimi K30.0000$19.0542.1M297k14h 45m2193

PostgreSQL 18 on SQLite

task page →4 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0714$227.56673.8M1.2M20h 0m15281
2Kimi K30.0659$265.37772.2M1.4M20h 1m14271
3Kimi K30.0604$238.20702.8M1.4M20h 1m15131
4Kimi K30.0495$197.52561.9M1.3M20h 0m11681

Quantum ESPRESSO pw.x in Rust

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.2602$75.00147.3M1.2M19h 51m4281
2Kimi K30.0488$160.13415.5M1.7M20h 0m9991
3Kimi K30.0244$100.42211.3M1.8M20h 0m5661
4Kimi K30.0081$131.87355.0M1.3M19h 56m7856
5Kimi K30.0000$72.76158.2M1.2M20h 1m4181

Qubit Routing

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.7526$40.44107.3M401k19h 37m38717
2Kimi K30.7151$55.15146.5M497k20h 0m48414
3Kimi K30.6987$79.07224.2M641k20h 1m55612
4Kimi K30.6270$30.7976.7M394k12h 20m31820
5Kimi K30.6254$97.66258.3M819k19h 58m57617

Reconnaissance Blind Chess Recovery

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.7571$29.2377.1M277k14h 20m3142
2Kimi K30.5071$31.5182.4M325k14h 46m3114
3Kimi K30.4286$27.4364.3M286k14h 50m25611
4Kimi K30.3143$28.9169.9M341k12h 50m2532
5Kimi K30.1429$43.61116.1M397k14h 40m3952

Fitness-Recap Video in Remotion

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.1767$191.75571.0M982k19h 53m13563
2Kimi K30.1723$171.83504.1M1.1M19h 44m12348
3Kimi K30.1716$266.42802.2M1.3M20h 1m16181
4Kimi K30.0657$243.01727.7M1.1M20h 1m16331
5Kimi K30.0536$202.63587.9M1.0M19h 57m147912

Snooker Prediction

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0000$126.53346.0M738k19h 31m7788
2Kimi K30.0000$140.38372.9M1.4M19h 51m92526
3Kimi K30.0000$122.14346.5M798k19h 47m7938
4Kimi K30.0000$117.31316.7M1.0M19h 47m8066
5Kimi K30.0000$105.21273.3M660k20h 0m5483

SPICE Circuit Simulator in Rust

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0000$179.57474.0M1.9M19h 42m12159
2Kimi K30.0000$215.15588.2M1.9M19h 56m14053
3Kimi K30.0000$199.08565.4M1.4M19h 58m12711
4Kimi K30.0000$204.94570.3M1.7M19h 34m13993
5Kimi K30.0000$193.03529.1M1.8M19h 58m12631

Synthetic Music Diarization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.4318$13.3633.2M179k18h 52m2445
2Kimi K30.4157$22.8160.7M202k19h 10m3477
3Kimi K30.4034$13.3431.5M191k18h 19m2474
4Kimi K30.3852$16.4839.9M237k19h 5m2342
5Kimi K30.3635$16.0039.1M227k18h 59m2346

Vision-only TORCS Racing Bot

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.0612$51.38138.7M388k19h 16m4569
2Kimi K30.0382$61.72163.7M377k19h 53m5343
3Kimi K30.0211$44.44114.5M382k20h 0m3861
4Kimi K30.0143$59.95154.4M389k20h 1m5072
5Kimi K30.0000$56.92159.4M376k20h 0m5211

† score adjusted by manual review; the trace's Judge tab carries the measured score and the rationale.

Verilog Simulator in Swift

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.8373$248.18704.4M1.6M20h 0m163611
2Kimi K30.7551$141.93387.6M1.1M20h 1m10091
3Kimi K30.7175$256.44747.8M1.5M20h 1m16692
4Kimi K30.6849$265.41774.0M1.6M19h 54m16822
5Kimi K30.6250$235.66685.5M1.4M19h 49m15211