Deep Research on DeepResearch Bench official 100-task-subset 1.0

0.5076RACE OverallOAgents-DR

Evaluation results

Method
OAgents-DR2026.01	0.5076	0.5039	0.512	0.5032	0.4941	0.3698	0.1256
WebWeaver2026.01	0.5062	0.5129	0.51	0.4998	0.4889	0.7825	1.6673
Gemini-2.5-Pro Deep Research2026.01	0.4888	0.4853	0.485	0.4918	0.4944	0.8144	1.1121
O-Researcher-RL2026.01	0.4848	0.4732	0.4954	0.4864	0.4758	0.3199	0.2601
O3 Deep Research2026.01	0.4698	0.4687	0.4525	0.4927	0.4714	0.7796	0.4079
GPT-52026.01	0.4677	0.4541	0.4454	0.5029	0.4747	0.3225	0.1221
O-Researcher-SFT2026.01	0.4624	0.4441	0.4684	0.4679	0.4676	0.2913	0.2263
MiniMax M22026.01	0.4606	0.4519	0.4456	0.4901	0.4468	0.1833	0.0288
Tongyi-Deep Research2026.01	0.4566	0.447	0.442	0.4882	0.442	—	—
Kimi-K2-Thinking2026.01	0.4565	0.4457	0.4394	0.4897	0.4465	0.1139	0.0097
MiroFlow2026.01	0.4487	0.4457	0.493	0.4545	0.4611	0.8698	0.2393
Kimi-K22026.01	0.4447	0.4278	0.3965	0.5082	0.46	0.0583	0.0027
OpenAI O32026.01	0.4371	0.4202	0.388	0.5029	0.459	0.19	0.011
Perplexity Deep Research2026.01	0.4225	0.4069	0.3939	0.464	0.4428	0.9024	0.3126
MiroThinker2026.01	0.4179	0.4022	0.3541	0.4779	0.4664	0.2632	0.0022
Claude-3-7-Sonnet2026.01	0.4067	0.3899	0.3766	0.4577	0.4146	0.9368	0.3248
Grok Deeper Search2026.01	0.4024	0.3797	0.3537	0.463	0.4405	0.8359	0.0815
Perplexity-Sonar-Reasoning-Pro2026.01	0.4022	0.3738	0.3611	0.4566	0.4474	0.3936	0.0835
Perplexity-Sonar-Pro2026.01	0.3893	0.3638	0.3426	0.447	0.4335	0.7866	0.1474
Gemini-2.5-Pro2026.01	0.3512	0.3406	0.2979	0.4167	0.3716	0.8181	0.3288
GPT-4.12026.01	0.3346	0.2942	0.2538	0.4233	0.4077	0.541	0.05
Qwen-2.5-72B-Instruct2026.01	0.3338	0.3027	0.2341	0.438	0.3949	0.4427	0.0896
Gemini-2.5-Flash2026.01	0.3239	0.3163	0.2673	0.3882	0.3448	0.8192	0.3108
Claude-3-5-Sonnet2026.01	0.2848	0.2482	0.2282	0.3512	0.3508	0.9404	0.0978