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 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0462$180.50154.3M515k19h 45m6315
2Grok 4.70.0329$52.4742.4M229k5h 56m21012
3Grok 4.70.0313$93.7878.9M256k12h 4m31314
4Grok 4.70.0218$33.8132.7M228k2h 52m1718
5Grok 4.70.0128$49.7744.3M205k7h 40m2316

Astronomy Toolkit

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.3076$28.4528.6M218k1h 15m1425
2Grok 4.70.2999$23.3123.1M224k1h 29m1183
3Grok 4.70.0108$18.1818.2M206k1h 1m1053
4Grok 4.70.0075$224.04218.5M550k3h 42m69915
5Grok 4.70.0032$13.3314.3M178k52m902

Cranelift Codegen Optimization

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0000$645.99652.2M2.1M20h 0m258915
2Grok 4.70.0000$228.14230.9M694k6h 44m9487
3Grok 4.70.0000$477.48483.7M1.5M16h 55m197520
4Grok 4.70.0000$635.61634.0M1.8M19h 42m254520
5Grok 4.70.0000$640.87669.7M1.8M19h 25m287727

Dart Style in Haskell

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.7672$810.72832.8M2.4M19h 51m3315144
2Grok 4.70.7101$934.33959.7M2.7M20h 0m378676
3Grok 4.70.6935$819.81827.3M2.4M19h 30m329279
4Grok 4.70.6906$879.91896.0M2.8M19h 59m356182
5Grok 4.70.4324$263.14264.6M900k6h 3m99620

FFmpeg libswscale Optimization

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0717$958.45972.7M3.5M19h 0m381051
2Grok 4.70.0692$75.9476.6M393k2h 13m3495
3Grok 4.70.0562$84.2786.0M423k2h 12m3535
4Grok 4.70.0482$1089.411092.6M3.0M15h 42m392949
5Grok 4.70.0022$90.4689.2M451k3h 23m3799

Crash-Proof Flash Filesystem

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0806$696.28696.9M2.0M12h 34m258417
2Grok 4.70.0184$312.58313.6M1.2M9h 0m128010
3Grok 4.70.0184$793.49801.3M2.8M20h 0m308423
4Grok 4.70.0173$406.03410.3M1.5M9h 50m162419
5Grok 4.70.0164$349.50343.9M1.4M11h 51m13498

FrogsGame Post-Training

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.4940$60.7936.4M219k19h 41m1691
2Grok 4.70.3640$259.38236.3M353k19h 57m7952
3Grok 4.70.2980$90.9963.2M209k18h 43m2772
4Grok 4.70.2920$118.9466.1M194k19h 26m2822
5Grok 4.70.0460$50.6836.5M192k18h 50m1962

Stepper Music Sequencer GBA

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.7472$628.06649.9M2.0M11h 13m261863
2Grok 4.70.6012$1385.341395.4M3.2M19h 29m506467
3Grok 4.70.4771$380.66393.3M1.3M7h 10m159732
4Grok 4.70.4712$1073.481101.7M3.6M19h 4m4513110
5Grok 4.70.4415$440.85450.4M1.7M8h 48m186319

Git to Zig

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.2530$1242.681266.7M4.0M20h 0m506870
2Grok 4.70.2469$910.55936.9M2.7M19h 58m376853
3Grok 4.70.2342$1107.041144.5M3.0M20h 0m463866
4Grok 4.70.2156$741.65767.0M2.0M19h 54m309973
5Grok 4.70.1876$623.38637.1M2.0M13h 32m260018

Granite Mamba2 Inference Optimization

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.8555$664.66670.3M2.4M20h 0m251555
2Grok 4.70.8303$663.28652.6M2.1M14h 34m234524
3Grok 4.70.7778$748.71746.1M3.0M20h 0m276323
4Grok 4.70.7473$45.1945.2M263k1h 21m1813
5Grok 4.70.4410$49.6548.3M283k2h 1m18413

Higgs Uncertainty Inference

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.7706$11.9313.0M204k1h 11m905
2Grok 4.70.0000$420.44408.4M739k19h 20m129318
3Grok 4.70.0000$7.848.3M179k58m633
4Grok 4.70.0000$17.2117.6M220k1h 13m995
5Grok 4.70.0000$9.2910.1M177k1h 2m762

SGLang Inference System Optimization

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0552$802.99801.5M2.3M19h 22m318324
2Grok 4.70.0000$211.95219.1M541k7h 14m8765
3Grok 4.70.0000$51.5950.1M157k1h 43m2125
4Grok 4.70.0000$100.20104.5M326k3h 50m4705
5Grok 4.70.0000$565.89566.8M1.5M11h 16m218910

Lean 4 Kernel Type Checker in Pascal

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.9589$192.69192.9M802k7h 19m7908
2Grok 4.70.9461$478.73472.8M1.7M19h 19m18875
3Grok 4.70.9103$351.92346.3M1.1M20h 0m14256
4Grok 4.70.8682$294.16294.8M1.1M13h 7m12296
5Grok 4.70.0000$418.16407.2M1.6M18h 42m16407

libexpat Optimization

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.3527$861.01886.1M3.3M19h 22m3633135
2Grok 4.70.1576$386.43383.9M880k5h 16m127713
3Grok 4.70.0000$77.7179.6M493k2h 16m3254
4Grok 4.70.0000$35.6933.2M254k1h 17m1395
5Grok 4.70.0000†$83.0379.6M516k2h 36m31413

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

Lua Native Compiler

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0000$74.2476.4M456k2h 25m3731
2Grok 4.70.0000$23.4623.7M161k55m1211
3Grok 4.70.0000$84.7885.6M413k2h 11m3303
4Grok 4.70.0000$10.1812.5M117k39m811
5Grok 4.70.0000$15.5217.3M144k46m1031

Machine-Learned Interatomic Potential

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.9608$60.4536.2M244k15h 59m1872
2Grok 4.70.9588$76.2457.2M274k16h 2m23812
3Grok 4.70.9549$72.0345.6M246k19h 9m2157
4Grok 4.70.9454$102.7267.6M289k18h 50m2548
5Grok 4.70.0000$87.6566.0M273k12h 6m26015

Medium-Range Weather Forecast

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.6200$170.45150.1M788k19h 40m61921
2Grok 4.70.5877$164.64141.5M437k18h 53m59629
3Grok 4.70.5339$192.06177.7M824k19h 10m75817
4Grok 4.70.4365$22.8623.1M216k4h 0m1303
5Grok 4.70.4219$34.8931.9M250k3h 58m13511

MEG Speech Decoding

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.1697$27.4625.8M199k3h 25m1335
2Grok 4.70.0000$17.0116.9M171k3h 5m1102
3Grok 4.70.0000$33.7429.5M210k5h 34m1575
4Grok 4.70.0000$9.8211.5M148k1h 49m852
5Grok 4.70.0000$32.9932.5M199k5h 31m1814

Wan 2.1 on MAX/Mojo

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.9226$177.41173.6M635k11h 45m7242
2Grok 4.70.9093$209.09202.2M818k13h 1m8442
3Grok 4.70.9053$77.0177.8M250k3h 37m3151
4Grok 4.70.0000$218.86185.5M568k19h 57m7918
5Grok 4.70.0000$85.9687.7M287k4h 37m3881

MS/MS De Novo Generation

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.1109$12.5612.3M170k1h 49m852
2Grok 4.70.1098$130.86121.5M481k15h 47m4984
3Grok 4.70.0968$55.0143.9M294k8h 31m1906
4Grok 4.70.0896$57.2942.5M256k5h 56m2036
5Grok 4.70.0000$47.0135.3M253k6h 31m1694

Multi-GPU Efficient Finetuning

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0000$60.2136.3M189k18h 34m1842
2Grok 4.70.0000$37.8924.1M164k18h 10m1402
3Grok 4.70.0000$29.8818.4M159k14h 59m1193
4Grok 4.70.0000$36.1819.8M164k20h 0m1265
5Grok 4.70.0000$59.6535.5M185k17h 44m1703

Notebook Compression

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.2584$211.20205.8M904k13h 1m85913
2Grok 4.70.2464$38.1737.7M250k3h 43m1928
3Grok 4.70.2422$40.6041.8M253k4h 6m22111
4Grok 4.70.2221$20.9623.3M213k3h 13m1495
5Grok 4.70.2015$41.0638.3M268k3h 17m1885

Flight-Sim Renderer in OpenGL

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.1336$487.10480.1M1.3M20h 1m183777
2Grok 4.70.1182$509.72506.8M1.5M18h 52m206885
3Grok 4.70.1138$588.60597.8M1.8M19h 55m2409106
4Grok 4.70.0713$540.56547.5M1.5M19h 14m2247102
5Grok 4.70.0196$518.13515.0M1.7M19h 40m207958

Optimizer Design

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.2867$182.37175.0M483k20h 0m74415
2Grok 4.70.2033$279.55260.9M639k20h 0m94917
3Grok 4.70.0119$55.6143.7M222k7h 51m1957
4Grok 4.70.0000$308.24304.6M520k20h 0m12151
5Grok 4.70.0000$72.6566.7M251k9h 7m2659

PostgreSQL 18 on SQLite

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.2747$697.58718.7M2.2M19h 51m297927
2Grok 4.70.2637$767.04780.4M2.4M18h 58m311730
3Grok 4.70.2527$961.79986.1M3.0M20h 0m390926
4Grok 4.70.1978$1004.111036.8M3.4M20h 0m421722
5Grok 4.70.0000$890.38903.6M3.1M20h 0m366012

Quantum ESPRESSO pw.x in Rust

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.6179$248.05251.2M1.1M11h 32m108616
2Grok 4.70.4553$566.99570.1M2.1M20h 0m22811
3Grok 4.70.2764$412.96420.2M1.8M14h 39m177917
4Grok 4.70.1138$593.38588.6M2.1M20h 0m23761
5Grok 4.70.0650$449.65416.6M1.1M20h 0m151711

Qubit Routing

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.7346$28.6227.0M233k2h 28m1197
2Grok 4.70.6849$35.2935.5M255k1h 35m15710
3Grok 4.70.6735$227.27226.0M738k19h 38m92575
4Grok 4.70.6585$21.7722.6M209k1h 19m1227
5Grok 4.70.6417$24.0124.2M210k1h 51m1314

Reconnaissance Blind Chess Recovery

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.8571$107.3195.1M427k11h 56m40026
2Grok 4.70.7821$110.9186.8M288k10h 45m30621
3Grok 4.70.7250$43.2641.6M230k3h 58m1883
4Grok 4.70.5893$172.69157.2M497k15h 0m62713
5Grok 4.70.5607$45.1844.4M217k3h 31m1978

Fitness-Recap Video in Remotion

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0961$439.29445.6M1.3M9h 16m1738103
2Grok 4.70.0840$675.93699.4M1.7M12h 52m279089
3Grok 4.70.0491$750.66786.8M2.0M15h 22m3238130
4Grok 4.70.0481$814.19833.8M1.7M19h 58m3225109
5Grok 4.70.0422$931.62946.8M1.9M20h 0m3560153

Snooker Prediction

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0086$81.1779.8M486k3h 8m3325
2Grok 4.70.0000$342.95338.0M1.2M19h 47m137039
3Grok 4.70.0000$282.03285.1M1.2M19h 51m119149
4Grok 4.70.0000$61.2564.2M387k4h 4m3355
5Grok 4.70.0000$79.8685.0M390k6h 4m38413

SPICE Circuit Simulator in Rust

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.7576$721.50746.3M2.8M15h 40m309416
2Grok 4.70.7172$780.09794.5M2.5M20h 0m328018
3Grok 4.70.7071$747.81772.9M3.1M17h 23m326717
4Grok 4.70.6970$807.64821.4M3.2M17h 39m336444
5Grok 4.70.6061$843.32872.9M3.2M19h 50m363616

Synthetic Music Diarization

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.4597$72.6761.2M249k8h 6m2494
2Grok 4.70.4458$66.8549.2M265k14h 13m20510
3Grok 4.70.4406$208.39177.1M399k19h 30m6124
4Grok 4.70.4284$43.8732.2M197k17h 17m1801
5Grok 4.70.4074$96.1879.9M265k13h 36m3078

Vision-only TORCS Racing Bot

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0456$436.54434.7M834k19h 58m173229
2Grok 4.70.0000†$129.68133.6M655k14h 50m60311
3Grok 4.70.0000†$87.5884.0M222k5h 40m3038
4Grok 4.70.0000†$51.7343.4M216k3h 56m1825
5Grok 4.70.0000†$353.91323.1M770k19h 55m118513

† 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 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.8853$200.79205.0M994k5h 44m8531
2Grok 4.70.8784$244.06246.8M1.1M15h 21m100527
3Grok 4.70.8596$839.06861.8M2.3M19h 58m350737
4Grok 4.70.8510$180.74176.1M878k5h 58m6584
5Grok 4.70.8476$334.11341.6M1.2M11h 19m145747