BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases7

DeepResearch Bench leaderboard

As of 6 Nov 2026, the top score on DeepResearch Bench is 55.3% by Claude Opus 4.6 (Anthropic), ahead of Claude Sonnet 4.6 at 54.9%, across 40 model configurations with a published result (DeepResearch Bench, result dated 5 Feb 2026). Research questions that require gathering sources and writing a report, judged by rubric. DeepResearch Bench is shown for reference only and does not count toward the BenchLeader Index.

Key facts (as of 6 Nov 2026)

  • Claude Opus 4.6 holds the top DeepResearch Bench score at 55.3% (DeepResearch Bench, 5 Feb 2026).
  • Claude Sonnet 4.6 is second on DeepResearch Bench at 54.9% (DeepResearch Bench, 17 Feb 2026).
  • DeepResearch Bench: Long-form research reports judged for depth and accuracy (DeepResearch Bench).
  • 40 model configurations have a published DeepResearch Bench result on BenchLeader (BenchLeader, 6 Nov 2026).
  • DeepResearch Bench does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).

Long-form research reports judged for depth and accuracy.

Published by
DeepResearch Benchdata via Epoch AI Benchmarking Hub
Category
Agents & tools
Index weight
Reference only
Models
40
Data as of
6 Nov 2026

CC BY 4.0 — Epoch AI, ‘AI Benchmarking Hub’, epoch.ai/benchmarks. Mirrored boards credit their original publishers.

What the test looks like

Research questions that require gathering sources and writing a report, judged by rubric.

How it is scored

Average rubric score, as published by DeepResearch Bench.

What to keep in mind

Rubric-graded reports; harness and search tools matter.

Top 15
  1. 1Claude Opus 4.6 (high)55.3%
  2. 2Claude Sonnet 4.6 (high)54.9%
  3. 3Claude Opus 4.5 (high)54.8%
  4. 4GPT-5.5 (high)54.0%
  5. 5Claude Opus 4.5 (low)53.7%
  6. 6Claude Opus 4.6 (medium)53.2%
  7. 7Claude Sonnet 4.552.6%
  8. 8Claude Opus 4.6 (low)51.4%
  9. 9Claude Sonnet 4.6 (low)50.4%
  10. 10Claude Opus 4.8 (high)50.2%
  11. 11Gemini 3 Flash (low)49.8%
  12. 12GPT-5.5 (medium)49.6%
  13. 13GPT-5 (low)49.6%
  14. 14Claude Opus 4.8 (low)49.3%
  15. 15Gemini 3 Flash (minimal)49.0%
40 of 40
#
1Claude Opus 4.6high55.3%62.62026-02-05
2Claude Sonnet 4.6high54.9%50.52026-02-17
3Claude Opus 4.5high54.8%59.02025-11-24
4GPT-5.5high54.0%63.72026-04-23
5Claude Opus 4.5low53.7%–2025-11-24
6Claude Opus 4.6medium53.2%–2026-02-05
7Claude Sonnet 4.552.6%53.72025-09-29
8Claude Opus 4.6low51.4%–2026-02-05
9Claude Sonnet 4.6low50.4%–2026-02-17
10Claude Opus 4.8high50.2%60.62026-05-28
11Gemini 3 Flashlow49.8%–2025-12-17
12GPT-5.5medium49.6%61.02026-04-23
13GPT-5low49.6%51.42025-08-07
14Claude Opus 4.8low49.3%–2026-05-28
15Gemini 3 Flashminimal49.0%53.92025-12-17
16GPT-5minimal48.9%45.32025-08-07
17GPT-5.5low48.7%58.22026-04-23
18GPT-5medium48.6%58.12025-08-07
19Claude Opus 4.148.3%51.42025-08-05
20GPT-5high48.1%57.92025-08-07
21Gemini 3 Flashhigh47.9%58.32025-12-17
22Gemini 3.1 Prohigh47.8%60.12026-02-19
23Claude Sonnet 4.5low47.5%–2025-09-29
24Claude Opus 4.8medium47.4%–2026-05-28
25Grok 447.3%58.32025-07-09
26Claude Opus 446.8%52.62025-05-22
27Claude Sonnet 446.6%50.52025-05-22
28Gemini 3 Prolow46.3%–2025-11-18
29Claude Haiku 4.5low45.5%–2025-10-15
30o3medium45.2%53.32025-04-16
31Gemini 3.1 Prolow44.5%–2026-02-19
32Claude 3.7 Sonnet43.6%49.92025-02-24
33Gemini 2.5 Pro42.8%54.52025-06-05
34GPT-5.1low42.8%–2025-11-13
35GPT-5.2low41.1%49.62025-12-11
36Gemini 3.1 Flash Litelow37.3%–2026-03-03
37Gemini 3.1 Flash Lite36.4%52.42026-03-03
38GPT-5.4 Minilow36.3%–2026-03-17
39GPT-5.4low35.1%58.92026-03-05
40DeepSeek R1 052835.1%49.32025-05-28

More agents & tools benchmarks · All LLM benchmarks

Related benchmarks

Top head-to-heads on DeepResearch Bench

Cite as: BenchLeader, “DeepResearch Bench leaderboard”, data as of 6 Nov 2026, https://www.benchleader.com/benchmarks/deepresearch_bench.

DeepResearch Bench: questions

What does DeepResearch Bench measure?
Research questions that require gathering sources and writing a report, judged by rubric. Scores are reported in percent of tasks solved; higher is better.
What is the top score on DeepResearch Bench?
As of 6 Nov 2026, the highest score on DeepResearch Bench is 55.3% by Claude Opus 4.6 (result dated 5 Feb 2026), ahead of Claude Sonnet 4.6 at 54.9%. Scores come from DeepResearch Bench and are re-read daily.
Which AI model leads DeepResearch Bench?
Claude Opus 4.6 leads DeepResearch Bench with 55.3% as of 6 Nov 2026 (result dated 5 Feb 2026), ahead of Claude Sonnet 4.6 at 54.9%.
How many models have DeepResearch Bench results?
40 model configurations have a DeepResearch Bench result on BenchLeader, all taken from DeepResearch Bench via Epoch AI Benchmarking Hub.
Who runs DeepResearch Bench and how often is it updated?
DeepResearch Bench is published by DeepResearch Bench. BenchLeader re-reads the published results once a day and records the date each result was published.
Does DeepResearch Bench count toward the BenchLeader Index?
No. DeepResearch Bench is shown for reference but left out of the composite index.