Search-SWE
CPU TASK Valid result

Reasoning-intensive Query Rewriting

GPT-6 Luna · Codex (v0.164.0) · reasoning effort max

Task score (%)30.0
Input Tokens19,348,336including 19,050,943 cached
Output Tokens188,140
Agent time2h 0min
Cost$0.32
Recorded steps135

Cost is estimated from recorded agent tokens at official API rates in USD (2026-11-01; DeepSeek off-peak pricing).

Loading trajectory…

Select this tab to browse the saved files.

Coding time & token usage
Original agent completion
agent_timeout
Coding started
2026-10-28T12:14:12.888647Z
Coding finished
2026-10-28T14:14:15.623804Z
Input (includes cache)
19,348,336
Cached input (subset)
19,050,943
Cache writes (subset)
221,771
Output
188,140
Reasoning (reported separately; not added)
137,296
Recorded responses
131

Input includes cached input; output is counted once. Auxiliary reasoning counts are shown as recorded. Regrading does not repeat coding usage.

Select this tab to browse the saved files.

Configuration

Evaluation settings summary
{
  "agent": "Codex",
  "agent_timeout_sec": 7200,
  "coding_commit": null,
  "effort": "max",
  "model": "openai/gpt-6-luna",
  "provider": "OpenRouter",
  "resources": null,
  "run_mode": "official_harbor_regrade",
  "task_commit": "93c1f4e3d69dd992a651853520a7a2b707e7851a",
  "verifier_commit": "93c1f4e3d69dd992a651853520a7a2b707e7851a",
  "version": "0.164.0"
}

Select this tab to browse the saved files.

Resolved configuration

Select this tab to browse the saved files.

Trial log

Select this tab to browse the saved files.

Exceptions

Select this tab to browse the saved files.

Select this tab to browse the saved files.

Trajectory downloads & provenance

Public text redacts private values and normalizes host paths. Source hashes identify the original records.

Trial ID
reasoning-query-rewriting-gpt-6-luna-max-cb1fbcd7bace
Source trajectory SHA-256
ab9e24c30a8c0b230af6e1210f9662640b986982a4c62ff4ec0a7c6b45be380c
Final reward SHA-256
5f77bd8c877abb8b58dc750d7290111f53edd9c060a393e0994ab40add4115f8