1496generated behavioral tests
87.9%official best score
82.0%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) | 20260517T094700Z |
Codex /goal | Mini-SWE-compatible no internet | 82.0% | ok | 1227/1496 | $1.95 | 34 | 0.16h | manifest · eval json · eval summary · usage audit |
| 2 | GPT 5.5 (xhigh) | 20260518T232006Z |
Codex /goal | Paper prompt + /goal no internet | 72.3% | ok | 1081/1496 | n/a | 33 | 0.15h | not exported |
| 3 | GPT 5.5 (xhigh) | 20260519T215046Z |
Codex /goal | Paper prompt + Goal contract no internet | 69.4% | ok | 1038/1496 | n/a | 70 | 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) · 20260517T094700Z
82.0% from 1227/1496 ProgramBench-scored tests. Raw public eval statuses: failure: 344, passed: 1483, skipped: 2.
Agent trace summary unavailable.
Likely miss class: behavioral mismatch.
84cbe4a71abetests.test_basic_invocation.test_help_short (failure)84cbe4a71abetests.test_basic_invocation.test_help_capital_h (failure)84cbe4a71abetests.test_benchmark.test_benchmark_default_level (failure)84cbe4a71abetests.test_benchmark.test_benchmark_range (failure)84cbe4a71abetests.test_error_handling.test_fast_negative_fails (failure)84cbe4a71abetests.test_error_handling.test_fast_zero_fails (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 | 87.9% | $8.10 | 84 |
| 2 | GPT 5.5 (high) | OpenAI | 82.7% | $2.92 | 42 |
| 3 | Claude Sonnet 4.6 | Anthropic | 82.7% | $25.44 | 501 |
| 4 | Claude Opus 4.7 (xhigh) | Anthropic | 80.6% | $4.56 | 100 |
| 5 | Claude Opus 4.7 | Anthropic | 80.0% | $5.56 | 155 |
| 6 | Claude Opus 4.6 | Anthropic | 78.7% | $9.86 | 216 |
| 7 | GPT 5.5 | OpenAI | 74.5% | $0.88 | 15 |
| 8 | GPT 5.4 | OpenAI | 46.6% | $0.24 | 10 |
| 9 | Gemini 3.1 Pro | 36.4% | $0.84 | 57 | |
| 10 | Claude Haiku 4.5 | Anthropic | 22.1% | $0.84 | 97 |
| 11 | GPT 5 mini | OpenAI | 14.4% | $0.01 | 9 |
| 12 | GPT 5.4 mini | OpenAI | 9.6% | $0.04 | 10 |
| 13 | Gemini 3 Flash | 5.2% | $0.20 | 46 |