Share your thoughts, 1 month free Claude Pro on usSee more

Long-form research on DRB

50.7Score

GPT-5 + Search

Updated 16d ago

Evaluation Results

Method	Links
GPT-5 + Search 2026.05		50.7
GPT-5 + Search 2025.11		50.7
GPT-5 + Our Search 2025.11		50.3
Gemini Deep Research 2026.05		48.8
Gemini Deep Research 2025.11		48.8
RubricEM-8B (RL, 1400 steps) 2026.05		47.8
OpenAI Deep Research 2026.05		46.9
OpenAI Deep Research 2025.11		46.9
Gemini 3 Pro + Search 2025.11		46.3
DR Tulu-8B (RL) 2025.11		45.4
Gemini 3.1 Pro + Search 2026.05		44.4
DR Tulu-8B (RL, 1900 steps) 2026.05		43.4
RubricEM-8B (SFT) 2026.05		43
Perplexity Deep Research 2026.05		42.3
Perplexity Deep Research 2025.11		42.3
Perplexity-Sonar (High) 2026.05		40.7
Perplexity-Sonar (High) 2025.11		40.7
WebThinker-32B-DPO 2026.05		40.6
Tongyi DeepResearch-30B-A3B 2026.05		40.6
Tongyi DeepResearch-30B-A3B 2025.11		40.6
WebThinker-32B-DPO (report) 2025.11		40.6
QwQ-32B 2025.11		40.3
DR Tulu-8B (SFT) 2026.05		39
DR Tulu-8B (SFT) 2025.11		39
WebThinker QwQ-32B 2026.05		37.9
WebThinker QwQ-32B (report) 2025.11		37.9
WebExplorer-8B 2026.05		36.7
WebExplorer-8B 2025.11		36.7
Ai2 ScholarQA – Claude Sonnet 2026.05		36.1
Ai2 ScholarQA - Claude Sonnet 2025.11		36.1
Claude-Sonnet Search 2026.05		34.5
Claude-Sonnet Search 2025.11		34.5
Qwen3-8B 2025.11		33.3
Qwen3-8B + Our Search 2026.05		28.2
WebThinker-32B-DPO 2025.11		23.3
Qwen3-8B + Our Search 2025.11		18.2
Search-R1-7B 2026.05		9.5
Search-R1-7B 2025.11		9.5
ASearcher-Web-7B 2025.11		7.8