Long-context text understanding on LongBench-Pro benchmark leaderboard · SOTA2 Research

Long-context text understanding on LongBench-Pro

37.7Overall ScoreFull attention

Evaluation Results

Method	Links
Full attention2026.05		37.7	35.88	38.96	31.13	33.72	31.6	48.86
UNIQUE2026.05		37.11	35.57	38.15	31.28	33.18	31.03	47.66
UNIQUE2026.05		36.72	35.77	37.98	31.02	32.78	30.75	48.34
UNIQUE2026.05		36.58	35.22	38.23	30.86	33.75	30.37	47.67
InfLLM2026.05		36.16	35.06	37.53	29.95	32.9	29.84	47.98
UNIQUE2026.05		35.91	35.11	36.88	30.23	33.38	29.11	46.94
InfLLM2026.05		35.42	34.77	36.68	29.21	32.58	28.6	48.12
InfLLM2026.05		34.99	33.82	36.21	28.96	32.38	27.31	46.74
UNIQUE2026.05		34.7	33.47	36.41	28.9	31.97	29	45.8
InfLLM2026.05		33.51	32.87	34.08	27.59	30.39	25.62	45.4
H2O2026.05		33.02	31.77	34.42	27.51	31.08	26.56	43.33
H2O2026.05		31.61	30.59	33.16	26.93	29.15	25.43	41.87
InfLLM2026.05		30.88	30.39	31.11	25.5	29.89	23.38	40.49
Quest2026.05		30.63	29.17	32.6	24.6	28.6	24.17	42.01
H2O2026.05		29.04	27.96	30.44	23.61	28.15	22.2	39.13
Quest2026.05		27.18	25.83	28.76	21.53	22.69	21.62	38.71
H2O2026.05		26.47	25.76	26.96	19.05	25.42	19.58	37.67
H2O2026.05		24.25	23.74	24.68	14.1	24.71	18.85	36.35
Quest2026.05		21.72	21.09	21.84	18.87	16.18	17.61	29.31
Quest2026.05		15.41	15.65	14.67	14.78	11.34	12.66	19.29
Quest2026.05		8.36	9.56	7.75	9.85	5.59	6.73	10.55