865generated behavioral tests
77.8%official best score
77.8%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) | 20260519T215046Z |
Codex /goal | Paper prompt + Goal contract no internet | 77.8% | ok | 673/865 | n/a | 62 | 0.22h | not exported |
| 2 | GPT 5.5 (xhigh) | 20260517T094700Z |
Codex /goal | Mini-SWE-compatible no internet | 68.0% | ok | 588/865 | $3.80 | 67 | 0.22h | manifest · eval json · eval summary · usage audit |
| 3 | GPT 5.5 (xhigh) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 43.2% | ok | 374/865 | n/a | 63 | 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) · 20260519T215046Z
Public eval evidence has not been exported for this row yet.
GPT 5.5 (xhigh) · 20260517T094700Z
68.0% from 588/865 ProgramBench-scored tests. Raw public eval statuses: failure: 404, passed: 1022, skipped: 2.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch.
6c819f915c59eval.tests.test_edge_cases.test_js_module_resolution (failure)6c819f915c59eval.tests.test_edge_cases.test_js_default_export (failure)6c819f915c59eval.tests.test_entropy_command.test_entropy_produces_valid_html (failure)6c819f915c59eval.tests.test_javascript_advanced.test_js_named_exports (failure)6c819f915c59eval.tests.test_javascript_advanced.test_js_namespace_import (failure)6c819f915c59eval.tests.test_javascript_advanced.test_js_re_export (failure)
GPT 5.5 (xhigh) · 20260518T232006Z
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 | 77.8% | $9.65 | 99 |
| 2 | Claude Sonnet 4.6 | Anthropic | 65.2% | $38.29 | 585 |
| 3 | GPT 5.4 | OpenAI | 62.5% | $0.30 | 8 |
| 4 | GPT 5.5 (high) | OpenAI | 59.1% | $4.51 | 55 |
| 5 | Claude Opus 4.6 | Anthropic | 56.8% | $12.17 | 245 |
| 6 | GPT 5.5 | OpenAI | 51.4% | $1.29 | 21 |
| 7 | Claude Opus 4.7 (xhigh) | Anthropic | 38.8% | $7.40 | 120 |
| 8 | Claude Opus 4.7 | Anthropic | 37.1% | $1.20 | 48 |
| 9 | Gemini 3 Flash | 35.0% | $0.32 | 96 | |
| 10 | Gemini 3.1 Pro | 18.0% | $1.75 | 105 | |
| 11 | GPT 5.4 mini | OpenAI | 10.9% | $0.03 | 9 |
| 12 | GPT 5 mini | OpenAI | 9.0% | $0.03 | 19 |
| 13 | Claude Haiku 4.5 | Anthropic | 0.0% | $0.86 | 154 |