structured output benchmark
Internal API instruction following (hard)
Internal API instruction following (hard) benchmark - specific documentation not found in official sources
Updated Aug 7, 2026
Published models7
Registry coverage7
MetricScore
EvidenceB
Internal API instruction following (hard) leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | OP | 64.0% | 100.0% | 7 | C | |
| 2 | OP | 54.0% | 83.3% | 7 | C | |
| 3 | OP | 50.0% | 66.7% | 7 | C | |
| 4 | OP | 49.1% | 50.0% | 7 | C | |
| 5 | OP | 45.1% | 33.3% | 7 | C | |
| 6 | OP | 31.6% | 16.7% | 7 | C | |
| 7 | OP | 29.2% | 0.0% | 7 | C |
Score distribution
Top published rows on the benchmark's original scale.
Internal API instruction following (hard) highlights
The top published results on this benchmark's own scale.
What is Internal API instruction following (hard)?
Definition and scoring fields from the benchmark registry.
Internal API instruction following (hard) benchmark - specific documentation not found in official sources
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- Internal API instruction following (hard)
- Modality
- text
- Primary category
- structured output
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- internal-api-instruction-following-(hard)|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about Internal API instruction following (hard).
Which model scores highest on Internal API instruction following (hard)?
GPT-5 is currently ranked first with 64.0%.
What does Internal API instruction following (hard) measure?
Internal API instruction following (hard) benchmark - specific documentation not found in official sources
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
7 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.