Is Claude Opus 4.6 good at Structured Output?
Claude Opus 4.6 ranks #30 of 43 for Structured Output — strong. The top pick for this task is qwen3.7-max-low.
Claude Opus 4.6 on each Structured Output sub-task
| Missing & Ambiguous Data | 90.0/100 | #34 |
| Schema Adherence | 90.0/100 | #23 |
| Transformation | 85.0/100 | #30 |
| Extraction | 80.0/100 | #28 |
| Noisy Structured Output Test | 69.2/100 | #40 |
Real examples, graded
WeakNormalize dates 55/100
“The model correctly extracted and normalized all dates to ISO 8601 format. However, it failed the strict JSON requirement by including markdown fences and trailing commentary, rendering the raw output unparseable as JSON.”
WeakClassify support tickets 55/100
“The model extracted the correct values and adhered to the enums, but failed the strict JSON requirement by including markdown fences and trailing commentary. Because the raw response cannot be parsed directly as JSON, it receives a 0 for schema validity and no wrapper text.”
WeakMissing field nulls 55/100
“The model extracted the data perfectly and used nulls exactly as requested. However, it failed the strict JSON requirement by including markdown fences and trailing commentary, rendering the raw output unparseable as JSON.”
Frequently asked
Is Claude Opus 4.6 good at Structured Output?
Claude Opus 4.6 ranks #30 of 43 models we tested for Structured Output, scoring strong.
What is Claude Opus 4.6's strongest Structured Output skill?
Its best sub-task here is Missing & Ambiguous Data.
This page is Spring Prompt, running
We just did this for every model. Do it for your prompt.
The rankings above come from running real tasks through real models and scoring every output. Spring Prompt is that same engine — pointed at your prompt, your test cases, and your definition of good.
- Generate test cases from your prompt — no eval set required to start.
- Compare models side by side with quality, cost and latency in one matrix.
- Optimise the winner until the scores say it's ready to ship.
Prompt × model results
12 test cases · 3 evals