StructEval text-format subset (StructEval-T)
We show this table for reference; we do not rank on it.
Measures whether a model produces valid structured output in text formats such as JSON, YAML and CSV.
Benchmark score on StructEval-T — October 29, 2026
We compile the StructEval-T rows from provider self-reports. The table contains 2 models. Nemotron 3.5 Bolt E4B BF16 has the highest published score at 80.40%, but the available coverage and evaluation setups do not establish a market leader. We do not use these results to rank models overall.
Nemotron 3.5 Bolt E4B BF16
NVIDIA
Nemotron 3.5 Bolt E4B
NVIDIA
2 modelsInstruction FollowingCurrentDisplay onlyUpdated October 29, 2026
Benchmark score table (2 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Nemotron 3.5 Bolt E4B BF16NVIDIA | 80.40% | Not reported | Open |
| 2 | Nemotron 3.5 Bolt E4BNVIDIA | 79.99% | Not reported | Open |
How to read this leaderboard
Operator receipt: 2 sourced rows are currently displayable on this page; the highest published score among these 2 models is Nemotron 3.5 Bolt E4B BF16 at 80.40%, which does not establish a market leader.
Honest limit: Covers the text-format subset only. Only NVIDIA-run results are stored so far, from one harness. Display-only and excluded from weighted ranking.
About StructEval-T
Year
2025
Tasks
Structured-output generation and conversion
Format
Average accuracy
Difficulty
Valid structured output
NVIDIA reports it for Nemotron 3.5 Bolt E4B (October 23, 2026), run in its NeMo Gym harness, as average accuracy on the text-only subset in reasoning mode. StructEval also has a renderable-format subset that is not stored here.
Freshness and provenance
Version
StructEval-T 2025
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does StructEval-T measure?
Measures whether a model produces valid structured output in text formats such as JSON, YAML and CSV.
Which model scores highest on StructEval-T?
Nemotron 3.5 Bolt E4B BF16 by NVIDIA currently leads with a score of 80.40% on StructEval-T.
How many models are evaluated on StructEval-T?
2 AI models have published results on StructEval-T in the BenchLM catalog.
Compare top models on StructEval-T
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.