913generated behavioral tests
98.4%official best score
88.9%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 | 88.9% | ok | 812/913 | n/a | 72 | 0.23h | not exported |
| 2 | GPT 5.5 (xhigh) | 20260517T094700Z |
Codex /goal | Mini-SWE-compatible no internet | 86.3% | ok | 788/913 | $5.51 | 112 | 0.28h | manifest · eval json · eval summary · usage audit |
| 3 | GPT 5.5 (xhigh) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 83.7% | ok | 764/913 | n/a | 80 | 0.21h | 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
86.3% from 788/913 ProgramBench-scored tests. Raw public eval statuses: failure: 127, passed: 836.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch, exact version/output mismatch.
fce1b7cdd1dbtests.test_angle_units.test_angle_conversion_deg (failure)fce1b7cdd1dbtests.test_arithmetic_operations.test_factorial (failure)fce1b7cdd1dbtests.test_arithmetic_operations.test_factorial_large (failure)fce1b7cdd1dbtests.test_arithmetic_operations.test_modulo (failure)fce1b7cdd1dbtests.test_edge_cases.test_modulo_with_decimals (failure)fce1b7cdd1dbtests.test_edge_cases.test_factorial_of_one (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 | 98.4% | $7.02 | 69 |
| 2 | GPT 5.5 (high) | OpenAI | 94.1% | $3.40 | 40 |
| 3 | Claude Opus 4.7 | Anthropic | 88.7% | $3.68 | 122 |
| 4 | Claude Opus 4.6 | Anthropic | 88.6% | $15.05 | 400 |
| 5 | Claude Sonnet 4.6 | Anthropic | 83.9% | $12.84 | 393 |
| 6 | GPT 5.4 | OpenAI | 75.8% | $0.33 | 13 |
| 7 | GPT 5.5 | OpenAI | 74.9% | $1.18 | 22 |
| 8 | Gemini 3 Flash | 66.9% | $0.26 | 116 | |
| 9 | GPT 5 mini | OpenAI | 60.5% | $0.04 | 23 |
| 10 | Claude Haiku 4.5 | Anthropic | 51.5% | $0.85 | 144 |
| 11 | GPT 5.4 mini | OpenAI | 44.9% | $0.06 | 8 |
| 12 | Claude Opus 4.7 (xhigh) | Anthropic | 0.5% | $22.33 | 294 |
| 13 | Gemini 3.1 Pro | 0.0% | $1.40 | 113 |