990generated behavioral tests
21.7%official best score
14.0%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) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 14.0% | ok | 139/990 | n/a | 51 | 0.15h | not exported |
| 2 | GPT 5.5 (xhigh) | 20260519T215046Z |
Codex /goal | Paper prompt + Goal contract no internet | 13.6% | ok | 135/990 | n/a | 49 | 0.16h | not exported |
| 3 | GPT 5.5 (xhigh) | 20260517T094700Z |
Codex /goal | Mini-SWE-compatible no internet | 11.6% | ok | 115/990 | $4.81 | 85 | 0.24h | manifest · eval json · eval summary · usage audit |
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) · 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.
GPT 5.5 (xhigh) · 20260517T094700Z
11.6% from 115/990 ProgramBench-scored tests. Raw public eval statuses: failure: 992, passed: 132, skipped: 48.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch.
1acdffe4f7e9tests.test_commands.test_append_to_existing_file (failure)1acdffe4f7e9tests.test_all_ids_vtabs.test_all_opids_concurrent_same_timestamp (failure)1acdffe4f7e9tests.test_db_display.test_union_query_display (failure)1acdffe4f7e9tests.test_all_ids_vtabs.test_all_opids_basic_access_log (failure)1acdffe4f7e9tests.test_field_overlay_display.test_field_display_mixed_log_formats (failure)1acdffe4f7e9tests.test_commands.test_tag_adds_to_logline (failure)
Official ProgramBench Results by Model
| # | Model | Provider | Score | Cost | Calls |
|---|---|---|---|---|---|
| 1 | GPT 5.5 (xhigh) | OpenAI | 21.7% | $11.63 | 87 |
| 2 | GPT 5.5 (high) | OpenAI | 20.9% | $3.57 | 39 |
| 3 | Claude Opus 4.6 | Anthropic | 13.4% | $10.16 | 341 |
| 4 | Gemini 3 Flash | 10.3% | $0.20 | 72 | |
| 5 | Claude Opus 4.7 | Anthropic | 8.8% | $1.07 | 43 |
| 6 | GPT 5.4 | OpenAI | 7.9% | $0.23 | 9 |
| 7 | GPT 5.5 | OpenAI | 7.2% | $0.83 | 13 |
| 8 | Claude Haiku 4.5 | Anthropic | 7.2% | $0.92 | 121 |
| 9 | Gemini 3.1 Pro | 5.7% | $0.91 | 57 | |
| 10 | GPT 5.4 mini | OpenAI | 2.4% | $0.03 | 11 |
| 11 | GPT 5 mini | OpenAI | 2.4% | $0.05 | 20 |
| 12 | Claude Opus 4.7 (xhigh) | Anthropic | 0.5% | $3.91 | 115 |
| 13 | Claude Sonnet 4.6 | Anthropic | 0.0% | $3.96 | 395 |