313generated behavioral tests
94.2%official best score
88.5%Codex best score
3Codex result rows
Codex Results by Model
| # | Model | Run | Agent | Mode | Score | Eval | Tests | Est. cost | Calls | Wall | Evidence |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | GPT 5.5 (xhigh) | 20260517T094700Z |
Codex /goal | Mini-SWE-compatible no internet | 88.5% | ok | 277/313 | $3.41 | 65 | 0.19h | manifest · eval json · eval summary · usage audit |
| 2 | GPT 5.5 (xhigh) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 88.2% | ok | 276/313 | n/a | 59 | 0.16h | not exported |
| 3 | GPT 5.5 (xhigh) | 20260519T215046Z |
Codex /goal | Paper prompt + Goal contract no internet | 80.8% | ok | 253/313 | n/a | 59 | 0.20h | not exported |
Why Scores Differ
Official ProgramBench rows are the public mini-SWE-agent submissions. GoalBench rows are separate Codex /goal submissions. Same model label does not mean the same agent, prompt, tool loop, or generated implementation. A GoalBench row is resolved only when the ProgramBench-scored pass rate is exactly 100%.
GPT 5.5 (xhigh) · 20260517T094700Z
88.5% from 277/313 ProgramBench-scored tests. Raw public eval statuses: failure: 36, passed: 302.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch.
edd50f76563beval.tests.test_map_mode.test_map_with_missing_source_files (failure)71456d35eecftests.test_errors.test_invalid_regex_unclosed_group (failure)71456d35eecftests.test_errors.test_invalid_map_not_json (failure)71456d35eecftests.test_errors.test_invalid_map_malformed_json (failure)71456d35eecftests.test_errors.test_conflict_without_overwrite_prepends_underscore (failure)71456d35eecftests.test_errors.test_no_source_option_provided (failure)
GPT 5.5 (xhigh) · 20260518T232006Z
Public eval evidence has not been exported for this row yet.
GPT 5.5 (xhigh) · 20260519T215046Z
Public eval evidence has not been exported for this row yet.
Official ProgramBench Results by Model
| # | Model | Provider | Score | Cost | Calls |
|---|---|---|---|---|---|
| 1 | GPT 5.5 (xhigh) | OpenAI | 94.2% | $5.15 | 70 |
| 2 | GPT 5.5 (high) | OpenAI | 85.0% | $2.16 | 35 |
| 3 | GPT 5.5 | OpenAI | 83.4% | $1.30 | 27 |
| 4 | Claude Sonnet 4.6 | Anthropic | 79.9% | $19.28 | 518 |
| 5 | Claude Opus 4.7 | Anthropic | 79.6% | $4.21 | 119 |
| 6 | Claude Opus 4.6 | Anthropic | 72.2% | $10.11 | 270 |
| 7 | Gemini 3 Flash | 61.0% | $0.29 | 142 | |
| 8 | Gemini 3.1 Pro | 53.7% | $1.54 | 170 | |
| 9 | GPT 5.4 | OpenAI | 51.4% | $0.30 | 15 |
| 10 | Claude Haiku 4.5 | Anthropic | 26.5% | $0.66 | 114 |
| 11 | GPT 5 mini | OpenAI | 15.0% | $0.02 | 11 |
| 12 | GPT 5.4 mini | OpenAI | 5.8% | $0.05 | 16 |
| 13 | Claude Opus 4.7 (xhigh) | Anthropic | 2.9% | $16.99 | 247 |