DRACO
100 deep-research tasks from real Perplexity queries across 10 domains, graded on about 40 expert rubric criteria each.
The makers built and saturation-tested it with Perplexity's own system. No public code repo found.
- Made by
- Perplexity
- Released
- February 2026
- Kind
- Deep research
- Level
- Expert
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Strongest model tested
- Claude Fable 5.2 #1 on the capability index
- Last in a launch post
- October 2026
- Last model added
- November 2026
In the authors’ words
We present DRACO (Deep Research Accuracy, Completeness, and Objectivity), a benchmark of complex deep research tasks. These tasks, which span 10 domains and draw on information sources from 40 countries, originate from anonymized real-world usage patterns within a large-scale deep research system. Tasks are sampled from a de-identified dataset of Perplexity Deep Research requests, then filtered and augmented to ensure that the tasks are anonymized, open-ended and complex, objectively evaluable, and representative of the broad scope of real-world deep research use cases. Outputs are graded against task-specific rubrics along four dimensions: factual accuracy (accuracy), breadth and depth of analysis (including completeness), presentation quality (including objectivity), and citation quality. DRACO is publicly available at https://hf.co/datasets/perplexity-ai/draco.
Claude Opus 5 leads with 88.6%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
- 1 Claude Opus 5 88.6%
- 2 Claude Fable 5.1 87.7%
- 3 Claude Fable 5.2 87.7%
- 4 Claude Opus 5.5 87.4%
- 5 Claude Sonnet 5.5 87%
- 6
Ling-3.1-flash 85.5% - 7
Step 5 Preview 83.3% - 8 Claude Haiku 5.5 81.5%
- 9 Claude Opus 4.8 80.4%
- 10 Claude Opus 4.7 77.7%
- 11
Tencent Hy4 preview 77.2% - 12
MiniMax-M3 73.2%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 5 | 88.6% | Jul 2026 | Lab model card | |
| 2 | Claude Fable 5.1 | 87.7% | Sep 2026 | Lab model card | |
| 3 | Claude Fable 5.2 | 87.7% | Nov 2026 | Lab model card | |
| 4 | Claude Opus 5.5 | 87.4% | Sep 2026 | Lab model card | |
| 5 | Claude Sonnet 5.5 | 87% | Sep 2026 | Lab model card | |
| 6 | Ling-3.1-flash | 85.5% | Oct 2026 | Lab model card | |
| 7 | Step 5 Preview | 83.3% | Sep 2026 | Lab model card | |
| 8 | Claude Haiku 5.5 | 81.5% | Oct 2026 | Lab model card | |
| 9 | Claude Opus 4.8 | 80.4% | May 2026 | Lab model card | |
| 10 | Claude Opus 4.7 | 77.7% | Apr 2026 | Lab model card | |
| 11 | Tencent Hy4 preview | 77.2% | Aug 2026 | Lab model card | |
| 12 | MiniMax-M3 | 73.2% | Jun 2026 | Lab model card |
3 models have held number one in labs’ own numbers since its release in February 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Labs’ model cards
Anthropic’s own number for Claude Opus 5 passed 85% in July 2026
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
4 labs put it in a launch post in the last six months
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Oct 13, 2026 | Ling-3.1-flash | 85.5 | Draco |
| Sep 20, 2026 | Step 5 Preview | 83.3 | Draco |
| Aug 28, 2026 | Tencent Hy4 preview | 77.2 | DRACO |
| Jun 1, 2026 | MiniMax-M3 | 73.2 | DRACO |
8 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Nov 9, 2026 | Claude Fable 5.2 system card | 87.7 | DRACO normalized score at max effort, 980k-token budget, Claude Opus 4.6 judge, five grading runs, 87.7% printed on chart. Judge and harness differ from the original paper (Figure 8.11.2.A) | Capabilities |
| Oct 7, 2026 | Claude Haiku 5.5 system card | 81.5 | DRACO, normalized score at max effort, Claude Opus 4.6 judge, printed on chart (Figure 8.8.2.A) | Capabilities |
| Sep 28, 2026 | Claude Sonnet 5.5 system card | 87 | DRACO normalized score at max effort, 980k token budget (8.11.2, Figure 8.11.2.A) | Capabilities |
| Sep 22, 2026 | Claude Opus 5.5 system card | 87.4 | DRACO normalized score at max effort, 980k-token budget, Claude Opus 4.6 judge, printed on chart (8.11.2, Figure 8.11.2.A) | Capabilities |
| Sep 1, 2026 | Claude Fable 5.1 system card | 87.7 | DRACO normalized score at max effort, 980k-token budget, Claude Opus 4.6 judge, printed on chart (Figure 8.12.2.A) | Capabilities |
| Jul 24, 2026 | Claude Opus 5 system card | 88.6 | DRACO normalized score at max effort, 980k-token budget, Claude Opus 4.6 judge, printed on chart (8.10.4, Figure 8.10.4.A) | Capabilities |
| May 28, 2026 | Claude Opus 4.8 system card | 80.4 | DRACO normalized score, max effort, 1M-token limit with compaction, Claude Opus 4.6 judge (8.10.4) | Capabilities |
| Apr 16, 2026 | Claude Opus 4.7 system card | 77.7 | DRACO normalized score, adaptive thinking at max effort, 1M token limit, Claude Opus 4.6 judge (8.8.4) | Capabilities |
More in Tool use
- AutomationBench In use Agent completes end-to-end business workflows inside a simulated company exposing dozens of REST APIs. Graded on resulting state.
Gemini 4 Carbon 53.9% - MCP Atlas In use Multi-step workflows over real MCP servers. Scores tool-use competency on realistic tasks. Gemini 4 Argon 90.7%
- BFCL V4 In use Models call functions and tools, including multi-turn agentic tasks with web search and memory. Checked against expected calls and outcomes.
Qwen3.7-Max 75% - τ³-bench Banking In use Agents act as banking customer-service reps over many turns with a simulated user, following policy. Pass^1 on the final database state. Qwen 3.8 Max 55.1%
- Toolathlon-Verified In use 108 long-horizon tasks across 32 apps through about 600 MCP tools. Execution-based checkers inspect final environment state. GLM-5.3-Flash 78.4%
- DeepSearchQA In use Questions that need multi-step search and synthesis. Final answers graded against references. Kimi K3 95%