reasoning benchmark
DRACO Leaderboard
DRACO is a deep research benchmark for evaluating an agent's ability to gather, synthesize, and reason over information to answer complex research questions.
Updated Sep 5, 2026
Models2
Model coverage2
MetricScore
EvidenceB
DRACO Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelTE | Score77.20% | Percentile100.00% | Participants2 | EvidenceC | Evaluated |
| Rank02 | ModelMI | Score73.23% | Percentile0.00% | Participants2 | EvidenceC | Evaluated |
DRACO Highlights
The leading models and scores on this benchmark.
DRACO Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for DRACO
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is DRACO?
What DRACO measures and how its scores work.
DRACO is a deep research benchmark in the reasoning category.
It measures an agent's ability to gather, synthesize, and reason over information to answer complex research questions, using the Score metric reported as a ratio based on the average of official rubric scores across all questions.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- DRACO
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about DRACO.
Which model scores highest on DRACO?
Hy4 preview is currently ranked first with 77.20%.
What are the top three models on DRACO?
The current leaders are Hy4 preview (77.20%) and MiniMax M3 (73.23%).
Which DRACO model has the lowest official input price?
MiniMax M3 has the lowest matched official input price at $0.30 input / $1.2 output per 1M tokens.
Which models are fastest among DRACO results?
The fastest matched records are MiniMax M3 (6.61 tok/s via MiniMax).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.