Traces

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

Astronomy Toolkit

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9999$11.6230.2M300k2h 33m1313
2Gemini 4 Argon0.8999$8.3227.3M281k1h 26m1172
3Gemini 4 Argon0.5135$11.8339.0M318k1h 48m1542
4Gemini 4 Argon0.0000$8.8535.0M322k1h 26m1372
5Gemini 4 Argon0.0000$27.0017.6M236k2h 6m932

Cranelift Codegen Optimization

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.4043$256.70491.0M1.5M20h 49m111624
2Gemini 4 Argon0.2201$298.10572.4M2.1M20h 43m125624
3Gemini 4 Argon0.1959$361.37454.4M1.3M17h 22m100227
4Gemini 4 Argon0.1238$407.04524.2M1.6M20h 47m125528
5Gemini 4 Argon0.0000$212.69506.0M1.7M20h 46m118427

Crash-Proof Flash Filesystem

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.7378$152.44168.7M1.1M7h 34m4293
2Gemini 4 Argon0.5467$54.25127.4M708k4h 23m3002
3Gemini 4 Argon0.2570$245.29429.7M1.7M13h 47m9603
4Gemini 4 Argon0.1980$156.36254.9M1.4M8h 56m6242
5Gemini 4 Argon0.1446$77.78221.0M1.0M7h 15m5322

Dart Style in Haskell

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9654$218.42430.1M2.0M15h 3m101021
2Gemini 4 Argon0.9568$288.60454.6M2.0M16h 22m109314
3Gemini 4 Argon0.9534$274.81581.3M2.0M17h 49m136815
4Gemini 4 Argon0.9529$147.73628.9M2.2M16h 46m147823
5Gemini 4 Argon0.9457$337.69476.2M2.1M17h 14m105518

FFmpeg libswscale Optimization

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.1572$160.32346.6M1.9M11h 31m76513
2Gemini 4 Argon0.0000$156.84296.0M1.6M9h 55m66811
3Gemini 4 Argon0.0000$106.17273.4M1.3M8h 28m60517
4Gemini 4 Argon0.0000$58.88167.7M1.1M5h 59m4057
5Gemini 4 Argon0.0000$190.93224.2M1.2M7h 43m5318

Fitness-Recap Video in Remotion

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.2451$233.36596.2M1.7M16h 49m126012
2Gemini 4 Argon0.2401$112.07365.5M1.1M10h 13m8605
3Gemini 4 Argon0.2185$165.03335.8M998k11h 8m8052
4Gemini 4 Argon0.2055$210.50290.4M915k10h 7m7172
5Gemini 4 Argon0.1980$159.82263.8M756k8h 13m6363

Flight-Sim Renderer in OpenGL

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9334$206.78382.0M1.3M15h 47m8775
2Gemini 4 Argon0.9080$207.90379.4M1.5M15h 34m85410
3Gemini 4 Argon0.8458$261.90442.3M1.8M19h 2m10566
4Gemini 4 Argon0.7148$130.35185.9M816k8h 32m4493
5Gemini 4 Argon0.7072$170.81258.5M1.2M8h 45m6733

FrogsGame Post-Training

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.8920$46.0449.7M235k14h 34m1725
2Gemini 4 Argon0.8020$82.9162.7M281k14h 57m1664
3Gemini 4 Argon0.6900$75.1061.9M250k17h 44m1914
4Gemini 4 Argon0.6040$30.5635.0M214k7h 42m1483
5Gemini 4 Argon0.5940$87.6965.7M221k8h 37m1753

Git to Zig

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.5447$362.36783.2M2.2M20h 48m175326
2Gemini 4 Argon0.5225$190.93727.6M2.5M20h 51m165723
3Gemini 4 Argon0.4697$330.68663.7M1.9M20h 55m156911
4Gemini 4 Argon0.4294$398.78647.4M2.0M20h 52m148726
5Gemini 4 Argon0.4167$333.71642.4M2.0M20h 59m140231

Granite Mamba2 Inference Optimization

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9104$105.47115.9M895k5h 40m3156
2Gemini 4 Argon0.9046$83.33130.1M792k5h 37m3593
3Gemini 4 Argon0.8993$145.24118.2M769k6h 54m3318
4Gemini 4 Argon0.8884$91.0370.1M500k3h 12m1953
5Gemini 4 Argon0.8832$36.1898.9M556k3h 28m2363

Higgs Uncertainty Inference

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.3435$96.8559.0M525k6h 34m1552
2Gemini 4 Argon0.3199$35.6242.7M359k5h 46m1553
3Gemini 4 Argon0.2379$42.74116.3M781k7h 14m2814
4Gemini 4 Argon0.2135$100.4483.2M842k7h 14m2394
5Gemini 4 Argon0.1190$36.0353.9M536k3h 17m1623

Kolmogorov Audio Compression

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.1051$267.43253.9M1.3M20h 43m65014
2Gemini 4 Argon0.0786$189.34268.5M1.6M19h 23m66114
3Gemini 4 Argon0.0759$254.33211.1M1.1M19h 34m46611
4Gemini 4 Argon0.0648$246.66202.6M896k20h 43m50810
5Gemini 4 Argon0.0474$174.59205.7M1.1M13h 22m47218

Lean 4 Kernel Type Checker in Pascal

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon1.0000$45.6847.8M303k2h 17m1352
2Gemini 4 Argon1.0000$81.6285.5M322k3h 41m2042
3Gemini 4 Argon1.0000$56.5692.0M304k3h 17m2032
4Gemini 4 Argon0.9807$13.2439.9M250k1h 35m1172
5Gemini 4 Argon0.9807$15.0249.2M292k2h 22m1452

libexpat Optimization

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.6081$51.02180.3M915k5h 20m3888
2Gemini 4 Argon0.5881$18.3871.0M446k2h 16m1609
3Gemini 4 Argon0.5870$86.35125.1M819k4h 40m3404
4Gemini 4 Argon0.5630$26.2474.9M691k2h 55m1989
5Gemini 4 Argon0.4643$17.1966.1M478k2h 17m1584

Lua Native Compiler

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon1.0000$73.2990.2M317k3h 26m2122
2Gemini 4 Argon0.9976$24.6660.9M275k2h 36m1372
3Gemini 4 Argon0.9976$34.7250.2M259k2h 22m1362
4Gemini 4 Argon0.9976$36.9359.8M371k3h 58m1542
5Gemini 4 Argon0.9976$21.7152.8M265k2h 10m1422

Machine-Learned Interatomic Potential

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9973$144.92105.5M303k19h 54m27310
2Gemini 4 Argon0.9944$61.2455.2M244k20h 19m18914
3Gemini 4 Argon0.9938$60.3445.2M281k20h 31m16713
4Gemini 4 Argon0.9931$56.1857.5M250k20h 47m20910
5Gemini 4 Argon0.9823$48.4349.9M216k20h 21m19920

Medium-Range Weather Forecast

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.5653$75.4253.6M408k7h 44m1629
2Gemini 4 Argon0.5347$151.2388.8M463k15h 22m23910
3Gemini 4 Argon0.4825$53.2850.9M579k8h 27m1594
4Gemini 4 Argon0.4704$122.3082.4M652k20h 41m19310
5Gemini 4 Argon0.0000$44.5935.1M340k6h 53m1445

MEG Speech Decoding

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.3815$127.6583.3M493k10h 58m2197
2Gemini 4 Argon0.3541$66.6565.7M370k10h 16m2086
3Gemini 4 Argon0.3496$73.6173.2M523k11h 46m2159
4Gemini 4 Argon0.3253$89.4063.7M382k14h 54m1957
5Gemini 4 Argon0.1941$62.2252.9M318k10h 43m1825

MS/MS De Novo Generation

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.1524$41.9158.1M431k9h 28m1825
2Gemini 4 Argon0.1303$43.3966.9M512k10h 28m1957
3Gemini 4 Argon0.1252$72.7447.3M391k5h 42m1646
4Gemini 4 Argon0.0000$51.8259.1M521k9h 41m1785
5Gemini 4 Argon0.0000$75.1654.4M420k6h 54m1856

Multi-GPU Efficient Finetuning

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.3095$59.0860.8M237k16h 56m1985
2Gemini 4 Argon0.2143$71.7765.5M249k18h 37m2008
3Gemini 4 Argon0.1905$119.5980.8M260k19h 19m2303
4Gemini 4 Argon0.1429$59.3568.3M219k18h 18m2077
5Gemini 4 Argon0.1190$99.1676.3M304k18h 34m2334

Notebook Compression

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.4592$350.30290.6M1.5M20h 36m69025
2Gemini 4 Argon0.4467$232.09321.0M1.6M20h 40m70628
3Gemini 4 Argon0.4112$223.42270.8M1.5M21h 0m63423
4Gemini 4 Argon0.3971$258.69226.7M1.1M20h 39m49816
5Gemini 4 Argon0.3901$194.88235.0M1.3M15h 31m56613

Optimizer Design

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.3700$138.1888.1M358k17h 59m2257
2Gemini 4 Argon0.3302$144.66126.0M552k18h 30m3276
3Gemini 4 Argon0.2842$147.5293.7M408k17h 44m2466
4Gemini 4 Argon0.2729$106.8362.5M361k17h 2m2006
5Gemini 4 Argon0.2637$128.39105.4M394k12h 7m2616

PostgreSQL 18 on SQLite

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.6923$41.6362.6M389k2h 48m1922
2Gemini 4 Argon0.6923$43.9265.5M430k2h 31m1922
3Gemini 4 Argon0.6923$26.3282.0M301k2h 32m2242
4Gemini 4 Argon0.6923$25.4364.1M365k2h 39m1862
5Gemini 4 Argon0.6484$34.6793.0M375k2h 50m2502

Quantum ESPRESSO pw.x in Rust

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9187$119.37259.0M1.1M10h 5m6451
2Gemini 4 Argon0.9106$303.04318.6M1.2M11h 58m6912
3Gemini 4 Argon0.9024$154.81373.7M1.6M12h 47m8872
4Gemini 4 Argon0.9024$388.42369.1M1.6M13h 39m8676
5Gemini 4 Argon0.8699$138.25365.2M1.5M10h 38m8562

Qubit Routing

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.8617$30.0862.2M527k4h 53m1764
2Gemini 4 Argon0.8488$43.6267.5M441k6h 4m1945
3Gemini 4 Argon0.8155$105.2973.3M505k9h 43m1778
4Gemini 4 Argon0.7986$85.4056.4M543k6h 53m1504
5Gemini 4 Argon0.7982$74.0164.4M466k6h 51m1817

Reconnaissance Blind Chess Recovery

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9286$102.1388.3M440k7h 8m2247
2Gemini 4 Argon0.9214$93.51158.0M759k9h 17m42613
3Gemini 4 Argon0.8964$116.54131.7M765k9h 49m3738
4Gemini 4 Argon0.8893$85.6194.9M445k9h 1m1996
5Gemini 4 Argon0.8679$74.4368.7M419k7h 35m1924

SGLang Inference System Optimization

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.4359$139.21309.9M955k8h 57m7466
2Gemini 4 Argon0.0000$275.58621.4M1.5M15h 16m139122
3Gemini 4 Argon0.0000$100.53344.7M936k8h 35m8288
4Gemini 4 Argon0.0000$197.19445.3M1.3M12h 41m10229
5Gemini 4 Argon0.0000$160.21500.1M1.5M13h 56m116920

Snooker Prediction

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.6242$58.63165.8M890k6h 8m4182
2Gemini 4 Argon0.5920$39.04128.3M687k6h 49m3573
3Gemini 4 Argon0.5407$63.97191.0M855k6h 41m4643
4Gemini 4 Argon0.4552$134.67137.5M744k5h 5m3723
5Gemini 4 Argon0.2214$64.59161.0M705k6h 32m4203

SPICE Circuit Simulator in Rust

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9394$221.62329.8M1.9M11h 7m7313
2Gemini 4 Argon0.9091$147.35234.0M1.0M8h 39m5036
3Gemini 4 Argon0.8889$270.20399.0M2.4M13h 19m8836
4Gemini 4 Argon0.8788$203.70185.4M882k6h 7m4263
5Gemini 4 Argon0.8283$51.57147.9M741k4h 35m4035

Stepper Music Sequencer GBA

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9654$544.95521.5M2.1M21h 0m11623
2Gemini 4 Argon0.9520$272.84638.9M3.5M20h 55m13824
3Gemini 4 Argon0.9258$445.36558.0M2.3M20h 59m12651
4Gemini 4 Argon0.9237$409.68593.6M2.6M20h 55m14457
5Gemini 4 Argon0.9165$187.49667.9M3.3M20h 59m15072

Synthetic Music Diarization

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.5622$95.7586.8M749k11h 34m2767
2Gemini 4 Argon0.5493$59.7663.4M476k9h 43m2015
3Gemini 4 Argon0.5456$46.2659.1M483k7h 58m1695
4Gemini 4 Argon0.5249$103.39100.4M544k9h 12m2468
5Gemini 4 Argon0.4215$32.5449.3M491k6h 13m1595

Verilog Simulator in Swift

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9349$95.9383.0M397k2h 56m1992
2Gemini 4 Argon0.9281$95.8972.5M365k2h 52m1783
3Gemini 4 Argon0.9281$83.8872.7M366k2h 46m1663
4Gemini 4 Argon0.9212$27.8068.8M403k2h 11m1654
5Gemini 4 Argon0.9178$56.5079.3M431k2h 53m1912

Vision-only TORCS Racing Bot

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.0000†$81.7158.0M243k8h 45m1803
2Gemini 4 Argon0.0000†$54.6248.6M203k6h 19m1692
3Gemini 4 Argon0.0000†$123.8083.2M286k8h 31m2043
4Gemini 4 Argon0.0000†$78.7163.3M288k10h 39m1885
5Gemini 4 Argon0.0000†$96.3987.4M314k10h 9m1983

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

Wan 2.1 on MAX/Mojo

task page →5 runs · indexed 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 4 Argon0.9182$55.0365.4M246k4h 0m1712
2Gemini 4 Argon0.9115$102.5378.3M305k5h 26m1812
3Gemini 4 Argon0.9073$129.59105.1M283k6h 53m2162
4Gemini 4 Argon0.9069$23.6061.0M217k3h 19m1752
5Gemini 4 Argon0.1758$81.7398.5M410k6h 12m2591