Emotional Support Dialogue on Emotional Support Dialogue Worst-case 1.0 (test) benchmark leaderboard · SOTA2 Research

Emotional Support Dialogue on Emotional Support Dialogue Worst-case 1.0 (test)

4.18Human-likenessGPT-5.4

Evaluation Results

Method	Links
GPT-5.42026.05		4.18	1.82	3.12	3.04	3.02	4.32	3.46	2.02	1.52	4.06	2.92	2.04	3.46	2.92
DS-V4-Pro2026.05		4.06	1.68	2.88	2.64	2.52	4.34	2.98	1.94	1.54	4.24	2.2	1.86	3.1	2.28
Qwen-3.62026.05		4.02	1.68	2.82	2.54	2.62	4.4	2.86	1.88	1.5	4.32	2.14	1.86	3	2.12
Claude-4.62026.05		3.96	1.7	2.76	2.52	2.4	4.22	2.8	1.82	1.38	4.08	2.2	1.92	3.1	2.28
DS-V4-Flash2026.05		3.64	1.58	2.48	2.24	2.06	4.16	2.24	1.74	1.3	3.9	2	1.82	2.96	2
Doubao2026.05		3.46	1.42	2.32	2.04	1.84	4.16	2.12	1.5	1.14	3.82	1.82	1.58	2.84	1.72
DS-V3.22026.05		3.2	1.36	2.28	2.16	1.84	4.04	2.06	1.62	1.1	3.76	2.02	1.88	3	1.9
GPT-4o2026.05		2.62	1.2	2.06	1.96	1.46	4.1	1.64	1.44	1.08	3.5	1.76	1.5	2.88	1.44
Qwen3-8B2026.05		2.42	1.24	2.08	1.96	1.22	4.06	1.36	1.46	1.12	3.54	1.76	1.54	2.88	1.38
EmoLLM2026.05		2.4	1.12	1.94	1.84	1.3	4	1.56	1.3	1	3.38	1.6	1.16	2.66	1.38
SoulChat2.02026.05		2.34	1.12	2.04	2	1.22	4.02	1.68	1.34	1.02	3.26	1.82	1.62	2.82	1.58
SimPsybot_q2026.05		2.26	1.14	1.88	1.92	1.24	4	1.66	1.32	1	3.02	1.5	1.72	2.76	1.6
Qwen3-4B2026.05		2.18	1.14	2.04	1.92	1.1	4.06	1.26	1.5	1.06	3.58	1.74	1.42	2.86	1.32
MindChat2026.05		2.04	1.1	1.46	1.4	1.14	3.92	1.74	1.12	1	2.68	1.16	1.18	2.32	1.12
PsyChat2026.05		2	1.12	1.62	1.56	1.14	3.96	1.42	1.22	1.06	2.98	1.18	1.12	2.36	1.08
MeChat2026.05		1.9	1.06	1.6	1.52	1	3.98	1.24	1.18	1.02	2.9	1.2	1.14	2.46	1.08
CPsyCounX2026.05		1.64	1.06	1.6	1.52	1	4.02	1.06	1.34	1	2.66	1.24	1.28	2.52	1.12