structured output benchmark
Internal API instruction following (hard) benchmark - specific documentation not found in official sources
Updated Aug 11, 2026
Higher score ranks better on this benchmark.
| 01 | OP | 64.0% | 100.0% | 7 | C | |
| 02 | OP | 54.0% | 83.3% | 7 | C | |
| 03 | OP | 50.0% | 66.7% | 7 | C | |
| 04 | OP | 49.1% | 50.0% | 7 | C | |
| 05 | OP | 45.1% | 33.3% | 7 | C | |
| 06 | OP | 31.6% | 16.7% | 7 | C | |
| 07 | OP | 29.2% | 0.0% | 7 | C |
A closer view of the leading scores on this benchmark.
The leading models and scores on this benchmark.
What Internal API instruction following (hard) measures and how its scores work.
Internal API instruction following (hard) benchmark - specific documentation not found in official sources
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
Common questions about Internal API instruction following (hard).
GPT-5 is currently ranked first with 64.0%.
Internal API instruction following (hard) benchmark - specific documentation not found in official sources
Yes. Higher values rank better for this benchmark.
7 model results are currently shown.
Yes. This benchmark can contribute to the current LLMBoard capability score.