Structured Output Benchmark: leaderboard
Structured-output benchmark measuring schema-constrained generation with value accuracy, faithfulness, JSON validity, path recall, type safety, and perfect-output rates.
Metric: Overall (%). Source: interfaze.ai. Saturation forecast: Estimated already saturated. 30 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 87 | #79 |
| 2 | Gemini 3.1 Pro (Preview) | 86.9 | #58 |
| 3 | GLM-5.1 | 86.6 | #97 |
| 4 | Claude Opus 4.7 | 86.4 | #49 |
| 5 | Claude Sonnet 5 | 86.2 | #57 |
| 6 | Qwen 3.5 35B A3B | 86.1 | #256 |
| 7 | GLM-4.7 | 86.1 | #192 |
| 8 | Gemini 2.5 Flash | 86 | #245 |
| 9 | GPT-5.5 | 86 | #31 |
| 10 | Qwen 3 235B A22B | 85.7 | #313 |
| 11 | Claude Sonnet 4.6 | 85.4 | #90 |
| 12 | Claude Opus 4.6 | 85.3 | #63 |
| 13 | DeepSeek V4 Pro | 85.3 | #102 |
| 14 | Kimi K2.6 | 85.3 | #104 |
| 15 | GPT-4.1 | 85 | #248 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=structured-output-benchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.