joule-bench/results_real.json
David Charlot eda1628ec3 joule-bench: Resolution x energy/cost coding-agent benchmark
A solver-agnostic harness that grades coding-agent task success with an
executable test oracle (tests fail before, pass after; pristine test files
restored to prevent gaming) and pairs it with measured energy (jouleclaw
IOReport, idle-subtracted, sub-window readings flagged unreliable), vendor-billed
dollars, and token-estimated cloud energy. Solvers: reference, noop, claude,
gemini, and cascade (the real joule_core engine via joule-coded). --passes shows
L0 cache amortization.

Headline (RESULTS.md): on a realistic operation mix, the cascade and frontier
claude both resolve 100%, but the cascade costs ~10x less per task -- mechanical
ops close locally at L1 for free, novel ops escalate to the same model once, and
repeats are free from the L0 cache.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
2026-06-22 07:18:36 -04:00

188 lines
No EOL
4.5 KiB
JSON

{
"energy_meter": "Measured (ModelBased, jouleclaw host counter, idle≈51.7 W subtracted)",
"frontier": [
{
"cloud_joules_per_resolved": 38445.9,
"cost_per_resolved": 0.237596125,
"joules_per_resolved": 0.0,
"median_wall_ms": 26208,
"provenance": "ModelBased",
"reliable_energy_runs": 3,
"resolution_pct": 66.66666666666667,
"resolved": 2,
"solver": "claude",
"tasks": 3,
"total_cloud_joules": 76891.8,
"total_cost_usd": 0.47519225,
"total_joules": 0.0
},
{
"cloud_joules_per_resolved": null,
"cost_per_resolved": null,
"joules_per_resolved": null,
"median_wall_ms": 0,
"provenance": "ModelBased",
"reliable_energy_runs": 0,
"resolution_pct": 0.0,
"resolved": 0,
"solver": "noop",
"tasks": 3,
"total_cloud_joules": 0.0,
"total_cost_usd": -0.0,
"total_joules": 12.069903
},
{
"cloud_joules_per_resolved": 0.0,
"cost_per_resolved": null,
"joules_per_resolved": 4.991112333333334,
"median_wall_ms": 1,
"provenance": "ModelBased",
"reliable_energy_runs": 0,
"resolution_pct": 100.0,
"resolved": 3,
"solver": "reference",
"tasks": 3,
"total_cloud_joules": 0.0,
"total_cost_usd": -0.0,
"total_joules": 14.973337
}
],
"runs": [
{
"baseline_failed": true,
"cost_usd": 0.2384065,
"energy_reliable": true,
"est_cloud_uj": 38813400000,
"input_tokens": 183977,
"microjoules": 0,
"note": "",
"output_tokens": 1009,
"provenance": "ModelBased",
"resolved": true,
"solver": "claude",
"task": "calc",
"wall_ms": 32111
},
{
"baseline_failed": true,
"cost_usd": 0.23678575,
"energy_reliable": true,
"est_cloud_uj": 38078400000,
"input_tokens": 179662,
"microjoules": 0,
"note": "",
"output_tokens": 1073,
"provenance": "ModelBased",
"resolved": true,
"solver": "claude",
"task": "fizzbuzz",
"wall_ms": 26208
},
{
"baseline_failed": true,
"cost_usd": 0.0,
"energy_reliable": true,
"est_cloud_uj": 0,
"input_tokens": 0,
"microjoules": 0,
"note": "",
"output_tokens": 0,
"provenance": "ModelBased",
"resolved": false,
"solver": "claude",
"task": "slicer",
"wall_ms": 1959
},
{
"baseline_failed": true,
"cost_usd": null,
"energy_reliable": false,
"est_cloud_uj": 0,
"input_tokens": null,
"microjoules": 5584226,
"note": "",
"output_tokens": null,
"provenance": "ModelBased",
"resolved": true,
"solver": "reference",
"task": "calc",
"wall_ms": 0
},
{
"baseline_failed": true,
"cost_usd": null,
"energy_reliable": false,
"est_cloud_uj": 0,
"input_tokens": null,
"microjoules": 5642953,
"note": "",
"output_tokens": null,
"provenance": "ModelBased",
"resolved": true,
"solver": "reference",
"task": "fizzbuzz",
"wall_ms": 1
},
{
"baseline_failed": true,
"cost_usd": null,
"energy_reliable": false,
"est_cloud_uj": 0,
"input_tokens": null,
"microjoules": 3746158,
"note": "",
"output_tokens": null,
"provenance": "ModelBased",
"resolved": true,
"solver": "reference",
"task": "slicer",
"wall_ms": 1
},
{
"baseline_failed": true,
"cost_usd": null,
"energy_reliable": false,
"est_cloud_uj": 0,
"input_tokens": null,
"microjoules": 4521968,
"note": "",
"output_tokens": null,
"provenance": "ModelBased",
"resolved": false,
"solver": "noop",
"task": "calc",
"wall_ms": 0
},
{
"baseline_failed": true,
"cost_usd": null,
"energy_reliable": false,
"est_cloud_uj": 0,
"input_tokens": null,
"microjoules": 3458960,
"note": "",
"output_tokens": null,
"provenance": "ModelBased",
"resolved": false,
"solver": "noop",
"task": "fizzbuzz",
"wall_ms": 0
},
{
"baseline_failed": true,
"cost_usd": null,
"energy_reliable": false,
"est_cloud_uj": 0,
"input_tokens": null,
"microjoules": 4088975,
"note": "",
"output_tokens": null,
"provenance": "ModelBased",
"resolved": false,
"solver": "noop",
"task": "slicer",
"wall_ms": 0
}
],
"suite": "tasks/smoke"
}