831generated behavioral tests
94.9%official best score
93.1%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 | 93.1% | ok | 774/831 | $4.74 | 91 | 0.22h | manifest · eval json · eval summary · usage audit |
| 2 | GPT 5.5 (xhigh) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 64.7% | ok | 538/831 | n/a | 85 | 0.24h | not exported |
| 3 | GPT 5.5 (xhigh) | 20260519T215046Z |
Codex /goal | Paper prompt + Goal contract no internet | 64.7% | ok | 538/831 | n/a | 110 | 0.29h | 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
93.1% from 774/831 ProgramBench-scored tests. Raw public eval statuses: failure: 60, passed: 876, skipped: 2.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch.
0dbd92599cb3tests.test_decompression.test_unpigz_decompresses_by_default (failure)eab1b8c8a035eval.tests.test_warnings.test_zlib_with_comment_warning (failure)efe9882e480atests.test_additional_flags.test_suffix_empty_rejected (failure)efe9882e480atests.test_additional_flags.test_T_flag_alternative_for_m (failure)efe9882e480atests.test_additional_flags.test_list_mode_zlib_format (failure)efe9882e480atests.test_decompression.test_cat_like_behavior_with_cdf_on_non_gzip (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 | 94.9% | $10.57 | 91 |
| 2 | Claude Opus 4.7 (xhigh) | Anthropic | 92.2% | $8.06 | 152 |
| 3 | GPT 5.5 (high) | OpenAI | 89.4% | $2.86 | 38 |
| 4 | GPT 5.5 | OpenAI | 84.2% | $1.03 | 17 |
| 5 | Claude Opus 4.7 | Anthropic | 83.2% | $4.53 | 149 |
| 6 | GPT 5.4 | OpenAI | 78.9% | $0.23 | 10 |
| 7 | Claude Sonnet 4.6 | Anthropic | 74.8% | $11.64 | 343 |
| 8 | Claude Haiku 4.5 | Anthropic | 73.4% | $0.47 | 84 |
| 9 | Gemini 3.1 Pro | 66.4% | $1.10 | 88 | |
| 10 | Gemini 3 Flash | 29.6% | $0.14 | 37 | |
| 11 | GPT 5.4 mini | OpenAI | 17.6% | $0.02 | 13 |
| 12 | GPT 5 mini | OpenAI | 10.6% | $0.01 | 8 |
| 13 | Claude Opus 4.6 | Anthropic | 0.0% | $1.63 | 215 |