Multi-modal Visualization Generation on MultiVis-Bench Basic Generation benchmark leaderboard · SOTA2 Research

Multi-modal Visualization Generation on MultiVis-Bench Basic Generation

69.97Low-Level Type ScoreMultiVis-Agent

Evaluation Results

Method	Links
MultiVis-Agent2026.01		69.97	42.57	37.62	76.24	92.08	77.23	65.95	87.95	87.43	74.5	68.65	80.12	92.64	80.02	72.99
MultiVis-Agent2026.01		66.01	36.3	19.8	73.6	84.16	72.94	58.8	82.01	82.87	70.79	59.57	69.88	90	73.44	66.12
Instructing LLM2026.01		54.58	43.46	22.55	69.61	71.57	66.01	54.63	75.08	75.88	64.46	62.83	65.6	79.41	68.7	61.66
LLM Workflow2026.01		52.81	34.98	6.27	47.85	39.27	61.06	40.37	64.85	65.08	57.1	49.67	52.23	70.63	57.99	49.18
Instructing LLM2026.01		50.5	40.92	12.54	58.75	40.26	56.77	43.29	64.93	66.53	58.25	51.49	55.53	71.78	59.46	51.37
LLM Workflow2026.01		50.5	37.29	14.85	50.5	55.45	67	45.93	75.91	78.12	64.93	60.07	64.36	79.54	68.52	57.23
nvAgent2026.01		27.72	15.18	3.3	49.5	59.74	48.18	33.94	33.33	30.73	30.78	20.13	33.75	49.54	31.43	32.68
nvAgent2026.01		24.42	16.83	2.97	45.21	53.8	42.24	30.91	30.94	29.41	26.49	21.95	31.68	46.67	29.69	30.3