641generated behavioral tests
96.1%official best score
75.5%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 | 75.5% | ok | 484/641 | $3.20 | 56 | 0.19h | manifest · eval json · eval summary · usage audit |
| 2 | GPT 5.5 (xhigh) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 74.3% | ok | 476/641 | n/a | 73 | 0.23h | not exported |
| 3 | GPT 5.5 (xhigh) | 20260519T215046Z |
Codex /goal | Paper prompt + Goal contract no internet | 73.2% | ok | 469/641 | n/a | 52 | 0.19h | 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
75.5% from 484/641 ProgramBench-scored tests. Raw public eval statuses: error: 30, failure: 354, passed: 563, skipped: 10.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch.
d1cae94adb26tests.test_colors.test_ls_colors_compound_extension (failure)d1cae94adb26tests.test_bytes.test_bytes_flag_with_hidden_and_all (failure)d1cae94adb26tests.test_colors.test_ls_colors_malformed_no_equals (failure)d1cae94adb26tests.test_core.test_basic_directory_traversal (failure)d1cae94adb26tests.test_colors.test_ls_colors_not_set (failure)d1cae94adb26tests.test_bytes.test_bytes_flag_exact_boundary_1024 (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 | 96.1% | $6.19 | 65 |
| 2 | GPT 5.5 (high) | OpenAI | 94.4% | $3.96 | 36 |
| 3 | Claude Opus 4.6 | Anthropic | 89.5% | $13.09 | 243 |
| 4 | Claude Opus 4.7 (xhigh) | Anthropic | 87.1% | $19.52 | 133 |
| 5 | GPT 5.5 | OpenAI | 79.1% | $1.14 | 17 |
| 6 | Claude Sonnet 4.6 | Anthropic | 78.3% | $17.62 | 335 |
| 7 | GPT 5.4 mini | OpenAI | 50.1% | $0.03 | 8 |
| 8 | GPT 5 mini | OpenAI | 44.8% | $0.01 | 8 |
| 9 | Claude Haiku 4.5 | Anthropic | 42.4% | $0.45 | 88 |
| 10 | Gemini 3 Flash | 38.7% | $0.33 | 81 | |
| 11 | GPT 5.4 | OpenAI | 36.5% | $0.14 | 8 |
| 12 | Claude Opus 4.7 | Anthropic | 0.3% | $19.26 | 164 |
| 13 | Gemini 3.1 Pro | 0.0% | $0.69 | 54 |