ResearchRubrics: A Benchmark of Prompts and Rubrics for Evaluating Deep Research Agents (ResearchRubrics)
We show this table for reference; we do not rank on it.
A deep-research agent benchmark that pairs realistic, domain-diverse prompts with more than 2,500 expert-written rubrics scoring factual grounding, reasoning soundness and clarity of long-form, evidence-backed answers.
Benchmark score on ResearchRubrics — November 11, 2026
We compile the ResearchRubrics rows from provider self-reports. LongCat-2.5 leads the table at 81.0%. We do not use these results to rank models overall.
1 modelAgenticCurrentDisplay onlyUpdated November 11, 2026
Benchmark score table (1 model)
The one result on this page was reported by the lab that built the model.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | LongCat-2.5MeituanReported by Meituan | 81.0% | Reported by Meituan | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
How to read this leaderboard
Operator receipt: 1 sourced row is currently displayable on this page; the leading published row is LongCat-2.5 at 81.0%.
Honest limit: Rubric grading uses an LLM judge, and launch rows are provider-run with harnesses and judges that differ by lab, so values are not directly comparable across sources.
About ResearchRubrics
Year
2025
Tasks
Deep-research prompts with 2,500+ expert-written rubrics
Format
Rubric compliance (%)
Difficulty
Open-ended deep research and long-form synthesis
Scale AI introduced ResearchRubrics in November 2025, built with more than 2,800 hours of human work, and released the prompts, rubrics and evaluation code. Scores are rubric compliance percentages; in the paper, Gemini Deep Research and OpenAI Deep Research stayed under 68%. Meituan reports LongCat-2.5 at 81.0 in its November 2026 model card, next to in-house runs of GPT-5.6 Sol (74.2), Claude Opus 5 (72.9), GLM-5.3 (79.9), Kimi K3 (78.8) and DeepSeek V4.1 Flash (80.9). BenchLM stores the LongCat-2.5 row; the other figures are Meituan runs listed for context.
Freshness and provenance
Version
ResearchRubrics 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 ResearchRubrics measure?
A deep-research agent benchmark that pairs realistic, domain-diverse prompts with more than 2,500 expert-written rubrics scoring factual grounding, reasoning soundness and clarity of long-form, evidence-backed answers.
Which model scores highest on ResearchRubrics?
LongCat-2.5 by Meituan currently leads with a score of 81.0% on ResearchRubrics.
How many models are evaluated on ResearchRubrics?
1 AI models have published results on ResearchRubrics in the BenchLM catalog.
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 7,000+ subscribers.
One email each week. Unsubscribe anytime.