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

Long-form research on ResearchQA

79.2Score

OpenAI Deep Research

Updated 2mo ago

Evaluation Results

Method	Links
OpenAI Deep Research 2026.05		79.2
GPT-5 + Search 2026.05		78.2
Perplexity Deep Research 2026.05		75.3
Ai2 ScholarQA – Claude Sonnet 2026.05		75
Gemini 3.1 Pro + Search 2026.05		74.5
RubricEM-8B (RL, 1400 steps) 2026.05		74.5
DR Tulu-8B (RL, 1900 steps) 2026.05		74.3
WebThinker-32B-DPO 2026.05		74.2
WebThinker QwQ-32B 2026.05		72.8
RubricEM-8B (SFT) 2026.05		71.8
Perplexity-Sonar (High) 2026.05		69.1
Gemini Deep Research 2026.05		68.5
DR Tulu-8B (SFT) 2026.05		68.5
Tongyi DeepResearch-30B-A3B 2026.05		66.7
WebExplorer-8B 2026.05		64.8
Claude-Sonnet Search 2026.05		64.3
Qwen3-8B + Our Search 2026.05		58.4
Search-R1-7B 2026.05		27.9