Vision & documents

OmniDocBench 1.5 leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, the highest measured score on OmniDocBench 1.5 is 92.1% by Qwen3.8 Max. 116 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Mythos 5	92.7%	estimated ± 7.4 pp, low confidence
2	Qwen3.8 Max	92.1%	measured
3	Claude Opus 5	92.0%	estimated ± 6.4 pp, low confidence
4	Claude Opus 5.5	92.0%	estimated ± 6.4 pp, low confidence
5	Gemini 3.6 Flash	92.0%	estimated ± 6.4 pp, low confidence
6	Gemini 3.7 Flash	92.0%	estimated ± 6.4 pp, low confidence
7	Gemini 3.8 Flash	92.0%	estimated ± 6.4 pp, low confidence
8	GPT-6.1 Sol	92.0%	estimated ± 6.4 pp, low confidence
9	GPT-6 Astra	92.0%	estimated ± 6.4 pp, low confidence
10	GPT-6 Sol	92.0%	estimated ± 6.4 pp, low confidence
11	Qwen3.8 Max Preview	92.0%	estimated ± 6.4 pp, low confidence
12	Grok 4.5	92.0%	estimated ± 6.4 pp, low confidence
13	Qwen3.8-Flash-Next	92.0%	estimated ± 6.4 pp, low confidence
14	GPT-6 Luna	92.0%	estimated ± 6.4 pp, low confidence
15	Apodex 1.1	92.0%	estimated ± 6.4 pp, low confidence
16	Apodex 1.1 Mini	92.0%	estimated ± 6.4 pp, low confidence
17	Gemini 3.5 Flash-Lite	92.0%	estimated ± 6.4 pp, low confidence
18	Ling 3.0 Flash VL	92.0%	estimated ± 6.4 pp, low confidence
19	Claude Opus 4.7 (Adaptive)	92.0%	estimated ± 6.4 pp, low confidence
20	Gemini 3 Flash	92.0%	estimated ± 6.4 pp, low confidence
21	GPT-5.3 Codex	92.0%	estimated ± 6.4 pp, low confidence
22	Claude Sonnet 5	92.0%	estimated ± 6.4 pp, low confidence
23	DeepSeek V4.1 Flash	91.9%	estimated ± 6.4 pp, low confidence
24	Claude Opus 4.7	91.7%	estimated ± 6.4 pp, low confidence
25	Mistral Large 4	91.7%	estimated ± 6.4 pp, low confidence
26	Gemini 3.1 Pro	91.7%	estimated ± 5.9 pp, low confidence
27	Gemini 3.5 Flash	91.7%	estimated ± 5.9 pp, low confidence
28	GPT-5.6 Sol	91.7%	estimated ± 5.9 pp, low confidence
29	Kimi K3	91.7%	estimated ± 5.9 pp, low confidence
30	Seed 2.1 Pro	91.7%	estimated ± 5.9 pp, low confidence
31	Gemini 3 Pro	91.7%	estimated ± 5.9 pp, low confidence
32	GPT-5.4	91.7%	estimated ± 5.9 pp, low confidence
33	GPT-5.5	91.7%	estimated ± 5.9 pp, low confidence
34	GPT-5.6 Terra	91.7%	estimated ± 5.9 pp, low confidence
35	Muse Spark	91.7%	estimated ± 5.9 pp, low confidence
36	Seed 2.1 Turbo	91.7%	estimated ± 5.9 pp, low confidence
37	GPT-5.2	91.7%	estimated ± 5.9 pp, low confidence
38	Kimi K2.6	91.7%	estimated ± 5.9 pp, low confidence
39	dots3-note Preview	91.7%	estimated ± 5.9 pp, low confidence
40	Qwen3.5 397B	91.7%	estimated ± 5.9 pp, low confidence
41	Qwen3.6 Plus	91.7%	estimated ± 5.9 pp, low confidence
42	Kimi K2.5	91.7%	estimated ± 5.9 pp, low confidence
43	Kimi K2.5 (Reasoning)	91.7%	estimated ± 5.9 pp, low confidence
44	GPT-5.6 Luna	91.7%	estimated ± 5.9 pp, low confidence
45	Grok 4.3	91.7%	estimated ± 5.9 pp, low confidence
46	Pareto 26.9	91.7%	estimated ± 5.9 pp, low confidence
47	MiMo-V2.5	91.7%	estimated ± 5.9 pp, low confidence
48	GPT-5.2-Codex	91.7%	estimated ± 6.4 pp, low confidence
49	Claude Opus 4.6	91.7%	estimated ± 5.9 pp, low confidence
50	Qwen3.8-Omni-Flash	91.6%	estimated ± 7.4 pp, low confidence
51	Gemma 4 31B	91.6%	estimated ± 5.9 pp, low confidence
52	MiniMax M3	91.6%	measured
53	GPT-5.4 mini	91.5%	estimated ± 5.9 pp, low confidence
54	Qwen3.7 Plus	91.4%	measured
55	Step 5 Preview	91.2%	estimated ± 5.9 pp, low confidence
56	Qwen3.8-27B	91.1%	measured
57	Qwen3.6-27B	91.0%	estimated ± 5.9 pp, low confidence
58	Claude Opus 4.8	90.8%	estimated ± 7.4 pp, low confidence
59	GLM-5.3-Flash	90.5%	estimated ± 7.4 pp, low confidence
60	GPT-5.1	90.3%	estimated ± 6.4 pp, low confidence
61	Claude Opus 4.6 (Adaptive)	90.0%	estimated ± 6.4 pp, low confidence
62	Muse Spark 1.1	89.9%	estimated ± 7.4 pp, low confidence
63	Qwen3.6-35B-A3B	89.9%	measured
64	Step 3.7 Flash	89.6%	estimated ± 6.4 pp, low confidence
65	Grok 4.20	89.5%	estimated ± 5.9 pp, low confidence
66	Sakana Fugu-Ultra	88.9%	estimated ± 7.4 pp, low confidence
67	Sakana Fugu	88.1%	estimated ± 7.4 pp, low confidence
68	Qwen3.5-122B-A10B	87.6%	estimated ± 6.4 pp, low confidence
69	Qwen3.5-27B	87.6%	estimated ± 6.4 pp, low confidence
70	Gemini 2.5 Pro	86.7%	estimated ± 6.4 pp, low confidence
71	Gemini 3.1 Flash-Lite	80.7%	estimated ± 7.4 pp, low confidence
72	GPT-5 (medium)	76.5%	estimated ± 6.4 pp, low confidence
73	Inkling-Small	75.8%	estimated ± 5.9 pp, low confidence
74	Muse Glimmer 30B	75.8%	measured
75	GPT-5 (high)	73.8%	estimated ± 6.4 pp, low confidence
76	Gemma 4 26B A4B	70.5%	estimated ± 5.9 pp, low confidence
77	Claude Opus 4.5 Thinking	67.2%	estimated ± 6.4 pp, low confidence
78	Inkling	60.5%	estimated ± 5.9 pp, low confidence
79	MiMo-V2.6-Flash	28.5%	estimated ± 6.4 pp, low confidence
80	GLM-5V-Turbo	18.2%	estimated ± 6.4 pp, low confidence
81	Qwen3.5-35B-A3B	15.4%	estimated ± 6.4 pp, low confidence
82	GPT-5.1-Codex	10.9%	estimated ± 6.4 pp, low confidence
83	GPT-5.1-Codex-Max	10.9%	estimated ± 6.4 pp, low confidence
84	Interfaze Beta	2.3%	estimated ± 5.9 pp, low confidence
85	Claude Opus 4.5	0.9%	estimated ± 5.9 pp, low confidence
86	Claude Sonnet 4.6	0.3%	estimated ± 6.4 pp, low confidence
87	o3	0.1%	estimated ± 6.4 pp, low confidence
88	MiMo-V2-Omni	0.1%	estimated ± 6.4 pp, low confidence
89	Gemma 4 12B	0.1%	estimated ± 5.9 pp, low confidence
90	Grok 4	0.0%	estimated ± 6.4 pp, low confidence
91	Claude 4.1 Opus Thinking	0.0%	estimated ± 6.4 pp, low confidence
92	GPT-5.4 nano	0.0%	estimated ± 5.9 pp, low confidence
93	Command A+	0.0%	estimated ± 5.9 pp, low confidence
94	LFM2.5-VL-3B	0.0%	estimated ± 5.9 pp, low confidence
95	Claude 3 Haiku	0.0%	estimated ± 6.4 pp, low confidence
96	Claude 4 Sonnet	0.0%	estimated ± 6.4 pp, low confidence
97	Gemini 1.5 Pro	0.0%	estimated ± 6.4 pp, low confidence
98	Gemini 2.5 Flash	0.0%	estimated ± 6.4 pp, low confidence
99	Gemma 3 27B	0.0%	estimated ± 6.4 pp, low confidence
100	Gemma 4 E2B	0.0%	estimated ± 6.4 pp, low confidence
101	Gemma 4 E4B	0.0%	estimated ± 6.4 pp, low confidence
102	GPT-4.1	0.0%	estimated ± 6.4 pp, low confidence
103	GPT-4.1 mini	0.0%	estimated ± 6.4 pp, low confidence
104	GPT-4.1 nano	0.0%	estimated ± 6.4 pp, low confidence
105	GPT-4o mini	0.0%	estimated ± 6.4 pp, low confidence
106	Grok 4.1 Fast	0.0%	estimated ± 6.4 pp, low confidence
107	Grok 4.1 Fast (Reasoning)	0.0%	estimated ± 6.4 pp, low confidence
108	Grok 4 Fast (Reasoning)	0.0%	estimated ± 6.4 pp, low confidence
109	LFM2.5-VL-1.6B-Extract	0.0%	estimated ± 6.4 pp, low confidence
110	Llama 4 Maverick	0.0%	estimated ± 6.4 pp, low confidence
111	Llama 4 Scout	0.0%	estimated ± 6.4 pp, low confidence
112	Mistral Large 3	0.0%	estimated ± 6.4 pp, low confidence
113	Mistral Medium 3	0.0%	estimated ± 6.4 pp, low confidence
114	Mistral Medium 3.5 128B	0.0%	estimated ± 6.4 pp, low confidence
115	Mistral Small 4	0.0%	estimated ± 6.4 pp, low confidence
116	Mistral Small 4 (Reasoning)	0.0%	estimated ± 6.4 pp, low confidence
117	Nemotron 3 Nano Omni 30B A3B	0.0%	estimated ± 6.4 pp, low confidence
118	Nova Pro	0.0%	estimated ± 6.4 pp, low confidence
119	Phi-4 Multimodal Instruct	0.0%	estimated ± 6.4 pp, low confidence
120	Qwen3.5 397B (Reasoning)	0.0%	estimated ± 6.4 pp, low confidence
121	Qwen3-Omni-30B-A3B-Instruct	0.0%	estimated ± 6.4 pp, low confidence
122	Qwen3-Omni-30B-A3B-Thinking	0.0%	estimated ± 6.4 pp, low confidence

OmniDocBench 1.5 leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General