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
1GPT-5.60.0416$183.22165.9M262k15h 47m47112
2GPT-5.60.0408$43.3247.2M136k19h 34m24910
3GPT-5.60.0224$49.4277.2M169k19h 53m42211
4GPT-5.60.0163$47.5455.3M185k19h 43m29110
5GPT-5.60.0159$41.6061.4M164k18h 50m36811

Astronomy Toolkit

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.4849$43.3852.6M186k3h 16m2635
2GPT-5.60.1823$39.2848.9M195k2h 7m2554
3GPT-5.60.0000$29.7437.3M170k2h 5m2118
4GPT-5.60.0000$23.2833.7M166k2h 33m2075
5GPT-5.60.0000$47.1154.0M192k2h 34m2624

Cranelift Codegen Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.0176$225.59218.6M216k6h 56m5066
2GPT-5.60.0000$297.21276.6M238k7h 17m5495
3GPT-5.60.0000$234.61222.5M240k7h 38m4615
4GPT-5.60.0000$212.19205.4M208k5h 52m4615
5GPT-5.60.0000$355.12344.3M254k10h 32m6273

Dart Style in Haskell

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.7949$998.14972.7M983k19h 18m214616
2GPT-5.60.6779$433.00423.6M391k6h 1m8444
3GPT-5.60.6344$89.8091.5M201k1h 55m3625
4GPT-5.60.6068$48.4752.9M204k1h 35m2743
5GPT-5.60.4547$71.7984.5M237k2h 25m4067

FFmpeg libswscale Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.0000$66.8275.1M179k1h 59m3188
2GPT-5.60.0000$78.4384.8M190k2h 4m3485
3GPT-5.60.0000$123.31124.9M233k2h 12m4223
4GPT-5.60.0000$76.6786.4M186k3h 5m36910
5GPT-5.60.0000$117.96120.8M233k3h 42m4058

Crash-Proof Flash Filesystem

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.0225$66.2368.5M180k1h 47m2748
2GPT-5.60.0225$81.1979.4M199k1h 55m2917
3GPT-5.60.0224$73.4072.5M171k1h 41m2738
4GPT-5.60.0173$159.04153.2M259k4h 31m44016
5GPT-5.60.0173$92.7287.0M212k1h 59m2687

FrogsGame Post-Training

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.0900$19.4628.7M100k17h 8m2062
2GPT-5.60.0540$20.8732.1M115k12h 56m2413
3GPT-5.60.0380$22.1134.6M103k18h 7m2232
4GPT-5.60.0360$61.3164.5M145k15h 2m2772
5GPT-5.60.0120$91.1851.7M86k17h 41m2284

Stepper Music Sequencer GBA

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.7701$944.54934.7M1.1M15h 1m241121
2GPT-5.60.7406$409.56413.9M542k8h 34m125115
3GPT-5.60.7219$253.31245.2M326k3h 19m6796
4GPT-5.60.6048$566.59550.5M752k14h 19m151311
5GPT-5.60.4981$490.17481.7M531k5h 24m127432

Git to Zig

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.3440$1238.081225.0M944k19h 8m262721
2GPT-5.60.2804$1082.721080.1M850k19h 36m231442
3GPT-5.60.2801$1175.971171.4M884k19h 37m260229
4GPT-5.60.2369$955.56945.7M722k19h 4m198213
5GPT-5.60.1874$770.31763.5M629k19h 0m16686

Granite Mamba2 Inference Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.7649$74.7474.2M207k2h 36m2705
2GPT-5.60.6303$111.93116.6M261k11h 2m4253
3GPT-5.60.6093$23.5332.7M162k1h 28m2073
4GPT-5.60.4658$119.15122.5M308k3h 17m4636
5GPT-5.60.3383$180.46175.3M266k4h 7m47410

Higgs Uncertainty Inference

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.7863$99.9097.4M228k4h 26m32720
2GPT-5.60.2084$52.2656.6M152k2h 42m2445
3GPT-5.60.0777$44.3350.5M161k1h 56m2296
4GPT-5.60.0080$56.4460.4M164k6h 2m25912
5GPT-5.60.0049$70.6365.2M124k2h 29m1946

SGLang Inference System Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.1092$291.23284.1M251k5h 12m61612
2GPT-5.60.0675$443.00437.3M281k6h 23m87714
3GPT-5.60.0168$283.36279.0M247k6h 47m65521
4GPT-5.60.0000$109.76114.5M139k5h 57m4017
5GPT-5.60.0000$182.18175.0M189k6h 4m4858

Lean 4 Kernel Type Checker in Pascal

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.9807$221.16222.7M429k6h 13m69119
2GPT-5.60.9791$78.0186.8M272k4h 32m3952
3GPT-5.60.9757$207.31199.1M381k7h 23m62010
4GPT-5.60.9339$340.15337.0M382k8h 35m78911
5GPT-5.60.7874$397.16404.0M396k8h 54m11347

libexpat Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.3847$146.13140.6M311k2h 25m4397
2GPT-5.60.3225$67.0675.3M223k1h 44m3269
3GPT-5.60.3025$18.3526.1M127k53m1444
4GPT-5.60.2868$49.1153.9M193k1h 32m2385
5GPT-5.60.2751$110.05104.7M243k2h 24m3507

Lua Native Compiler

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.0000$35.0744.0M172k2h 41m21110
2GPT-5.60.0000$7.458.9M65k25m653
3GPT-5.60.0000$127.72124.4M217k2h 5m38224
4GPT-5.60.0000$24.8437.1M163k1h 9m21913
5GPT-5.60.0000$20.5039.9M168k1h 9m2317

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

Machine-Learned Interatomic Potential

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.9925$108.64109.2M151k11h 16m3488
2GPT-5.60.9918$109.29114.5M152k15h 47m40416
3GPT-5.60.9437$26.2041.3M108k14h 19m2373
4GPT-5.60.0076$48.9956.5M122k14h 29m2935
5GPT-5.60.0010$36.5549.8M109k7h 20m2723

Medium-Range Weather Forecast

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.5936$28.8247.8M113k15h 13m2944
2GPT-5.60.0000$122.35137.1M209k17h 41m5767
3GPT-5.60.0000$63.6270.4M130k17h 56m2903
4GPT-5.60.0000$72.5794.1M189k15h 53m47814
5GPT-5.60.0000$88.96100.2M138k14h 52m4168

MEG Speech Decoding

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.3119$41.2143.5M133k5h 23m2042
2GPT-5.60.3039$14.1019.5M85k2h 25m1372
3GPT-5.60.2311$153.75128.1M180k10h 45m29817
4GPT-5.60.1902$64.3567.7M163k7h 2m2794
5GPT-5.60.1862$104.8098.3M165k7h 53m3188

Wan 2.1 on MAX/Mojo

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.9241$103.11107.4M168k5h 28m3698
2GPT-5.60.9087$34.0444.6M108k4h 5m1999
3GPT-5.60.9062$69.5278.1M124k3h 50m2953
4GPT-5.60.9014$57.1166.4M129k4h 8m2675
5GPT-5.60.8681$66.0672.8M105k4h 8m2674

MS/MS De Novo Generation

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.1316$55.6569.9M170k12h 29m3322
2GPT-5.60.1312$87.71101.7M199k15h 19m4536
3GPT-5.60.0000$29.4443.7M172k11h 11m2914
4GPT-5.60.0000$115.28137.3M233k11h 7m60310
5GPT-5.60.0000$33.4341.9M177k7h 37m2574

Multi-GPU Efficient Finetuning

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.2381$19.3323.5M110k15h 47m1673
2GPT-5.60.0238$19.6525.0M71k12h 53m1333
3GPT-5.60.0000$16.4525.9M78k16h 50m1802
4GPT-5.60.0000$24.1334.7M95k16h 10m1812
5GPT-5.60.0000$30.4139.1M90k14h 8m1993

Notebook Compression

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.2623$123.00167.4M279k11h 19m6288
2GPT-5.60.2599$99.01115.0M257k15h 3m52810
3GPT-5.60.2438$58.4876.1M190k17h 55m41214
4GPT-5.60.2326$88.19110.8M211k11h 35m5186
5GPT-5.60.1417$64.2780.6M214k10h 2m3893

Flight-Sim Renderer in OpenGL

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.2171$270.86274.0M362k12h 30m8064
2GPT-5.60.1552$336.67340.6M347k11h 28m90614
3GPT-5.60.1222$253.61257.1M305k8h 28m7603
4GPT-5.60.1119$309.74312.3M342k9h 22m85712
5GPT-5.60.1032$420.32406.2M354k11h 35m88013

Optimizer Design

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.3487$65.5380.8M133k10h 22m4334
2GPT-5.60.3194$196.66203.2M202k11h 49m67823
3GPT-5.60.2846$55.2972.7M133k15h 42m4235
4GPT-5.60.2809$71.6593.2M177k14h 22m52321
5GPT-5.60.2234$31.2952.9M100k13h 17m3732

PostgreSQL 18 on SQLite

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.5275$209.42198.9M277k4h 5m58912
2GPT-5.60.3901$654.39640.1M687k19h 58m147228
3GPT-5.60.0055$293.05285.8M311k4h 20m6716
4GPT-5.60.0000$137.59141.5M243k3h 27m4802
5GPT-5.60.0000$142.29142.1M259k4h 34m46211

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

Quantum ESPRESSO pw.x in Rust

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.5203$304.21296.3M260k4h 50m6069
2GPT-5.60.4634$57.4264.9M189k2h 47m27913
3GPT-5.60.4553$120.99116.9M214k5h 5m4048
4GPT-5.60.4065$104.97107.5M204k3h 16m39110
5GPT-5.60.0000$210.32203.5M250k6h 57m5776

Qubit Routing

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.7946$58.7777.9M203k6h 45m4238
2GPT-5.60.7781$75.1186.9M202k10h 41m41011
3GPT-5.60.7314$114.07131.6M228k8h 56m56914
4GPT-5.60.7190$154.11169.8M283k6h 40m68813
5GPT-5.60.7125$69.3392.1M185k8h 7m4755

Reconnaissance Blind Chess Recovery

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.8893$28.4245.1M149k6h 33m2917
2GPT-5.60.8607$35.9957.3M138k6h 13m33613
3GPT-5.60.8071$20.6232.5M99k10h 50m19610
4GPT-5.60.8000$20.5330.5M123k2h 52m1973
5GPT-5.60.7357$15.5121.1M106k11h 38m1689

Fitness-Recap Video in Remotion

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.2152$950.51938.4M1.0M19h 3m234345
2GPT-5.60.1930$656.93636.3M650k19h 12m147811
3GPT-5.60.1913$298.28288.7M261k7h 23m65112
4GPT-5.60.1895$283.99275.5M274k5h 43m6876
5GPT-5.60.1374$190.12189.9M230k5h 21m5892

Snooker Prediction

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.4857$119.18111.7M242k2h 5m29911
2GPT-5.60.4839$170.16165.0M275k4h 12m4329
3GPT-5.60.4428$157.87150.8M208k2h 34m3549
4GPT-5.60.4295$132.32126.4M240k4h 1m3696
5GPT-5.60.2890$179.27154.6M213k9h 23m3575

SPICE Circuit Simulator in Rust

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.4444$366.07350.2M455k6h 25m79722
2GPT-5.60.4141$307.32296.1M386k4h 30m78315
3GPT-5.60.3939$538.95528.2M617k6h 30m128021
4GPT-5.60.3333$187.27186.7M262k2h 15m5158
5GPT-5.60.2626$226.84221.2M327k3h 7m59213

Synthetic Music Diarization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.5247$44.1953.3M154k7h 34m2673
2GPT-5.60.5065$80.6578.5M178k15h 57m33911
3GPT-5.60.4952$59.6269.7M147k15h 37m3593
4GPT-5.60.4737$78.7876.5M133k14h 47m2864
5GPT-5.60.4727$68.5982.8M153k14h 29m4257

Vision-only TORCS Racing Bot

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GPT-5.60.1207$165.68160.5M163k9h 38m4167
2GPT-5.60.0000$58.1463.0M111k13h 6m2884
3GPT-5.60.0000$25.7538.4M104k16h 16m2184
4GPT-5.60.0000$51.9957.9M124k4h 17m2776
5GPT-5.60.0000$79.4380.8M113k10h 50m2854

† 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
1GPT-5.60.8853$427.16420.2M394k14h 37m9219
2GPT-5.60.8596$271.89260.9M315k4h 56m6505
3GPT-5.60.8493$335.11326.9M349k4h 23m7469
4GPT-5.60.8339$184.09180.1M273k4h 24m4806
5GPT-5.60.8236$181.11176.3M277k4h 52m4763