Share your thoughts, 1 month free Claude Pro on usSee more

Information Visual Question Answering on InfoVQA (test)

89.3ANLS

Seed-1.5-VL

Updated 22d ago

Evaluation Results

Method
Seed-1.5-VL 2026.03	89.3	-
Pixelis (Qwen3-VL-8B-Instruct) 2026.03	87.9	-
Pixelis: SFT + RFT 2026.03	86.7	-
PRM (process reward; tools; 8B) 2026.03	86.2	-
Pixelis: SFT + TTRL 2026.03	86.1	-
Pixel Reasoner (Qwen3-VL-8B-Instruct) 2026.03	85.8	-
Pixelis: RFT + TTRL 2026.03	85.8	-
Late-fusion 2026.03	85.6	-
Step Self-Consistency (step-level) 2026.03	85.4	-
Pixelis: SFT only 2026.03	84.9	-
Qwen-VL-Max-0809 2024.10	84.5	-
Qwen2-VL-72B 2024.12	84.5	-
Pixelis: RFT only 2026.03	84.4	-
InternVL2.5-78B 2024.12	84.1	-
RA-TTA (retrieval-augmented) 2026.03	84	-
InternVL2.5-38B 2024.12	83.6	-
RV Self-Consistency (answer-only) 2026.03	83.6	-
Pixelis: TTRL only 2026.03	83.5	-
Qwen3-VL-8B-Instruct 2026.03	83.1	-
Realistic TTA of VLMs (StatA) 2026.03	82.7	-
InternVL2-Llama3-76B 2024.12	82	-
Qwen3-VL-30B-A3B-Instruct 2026.03	82	-
Molmo-72B 2024.12	81.9	-
Gemini-1.5-Pro 2024.09	81	-
Gemini-1.5-Pro 2024.12	81	-
SmoLA PaLI-X 2024.02	80.3	-
InternVL2.5-26B 2024.12	79.8	-
GPT-4o-20240513 2024.12	79.2	-
InternVL2-40B 2024.12	78.7	-
DocCogito 2026.03	78.6	-
InternVL2.5-8B 2024.12	77.6	-
Qwen2-VL-7B 2024.12	76.5	-
DocCogito 2026.03	76	-
InternVL2-26B 2024.12	75.9	-
InternVL-1.5-Plus 2024.11	75.7	-
Qwen3-VL-Instruct 2026.03	75.7	-
Marten 2026.03	75.2	-
GPT-4V 2024.12	75.1	-
LLaVA-OneVision-72B 2024.12	74.9	-
InternVL2-8B 2024.12	74.8	-
InternVL2 2026.03	74.8	-
Qwen3-VL-Instruct 2026.03	74.6	-
Claude-3.5-Sonnet 2024.12	74.3	-
Molmo-7B-D 2024.12	72.6	-
InternVL-Chat-V1.5 2024.12	72.5	-
InternVL2.5-4B 2024.12	72.1	-
TextHawk2 2026.03	67.8	-
MM1.5-30B 2024.09	67.3	-
InternVL2-4B 2024.12	67	-
SMoLA-PaLI-X_FT (Specialist) 2023.12	66.2	-
ScreenAI 2024.02	65.9	-
SMoLA-PaLI-XFT 2023.12	65.6	-
Qwen2-VL-2B 2024.12	65.5	-
PaLI-3 (Specialist) 2023.12	62.4	-
SOTA [8] 2023.12	62.4	-
PaLI-3 2024.02	62.4	-
ScreenAI 2024.02	61.4	-
InternVL2.5-2B 2024.12	60.9	-
Mini-Monkey 2026.03	60.1	-
MM1.5-7B 2024.09	59.5	-
InternVL2-2B 2024.09	58.9	-
InternVL2-2B 2024.12	58.9	-
MM1.5-3B 2024.09	58.5	-
DocLayLLM 2026.03	58.4	-
Aquila-VL-2B 2024.12	58.3	-
PaLI-3 (Specialist) 2023.12	57.8	-
PaLI-3 2024.02	57.8	-
SMoLA-PaLI-3FT 2023.12	57.3	-
InternVL2.5-1B 2024.12	56	-
MM1.5-1B-MOE 2024.09	55.9	-
Claude-3-Opus 2024.12	55.6	-
PaLI-X 2023.12	54.8	-
Gemini Nano-2 2024.09	54.5	-
MM1.5-3B-MOE 2024.09	53.6	-
SMoLA-PaLI-3FT 2023.12	52.4	-
Gemini Nano-1 2024.09	51.1	-
InternVL2-1B 2024.12	50.9	-
DocOwl-1.5-Chat 2024.09	50.7	-
DocOwl-1.5-Chat 2026.03	50.7	-
MM1.5-1B 2024.09	50.5	-
DocOwl-1.5 2026.03	50.4	-
Zhang et al. 2026.03	50.2	-
SMoLA-PaLI-XFT 2023.12	49.2	-
Phi-3-Vision-4B 2024.09	49	-
DocFormerv2large 2023.06	48.8	-
DocFormerv2 2026.03	48.8	-
UDOP 2023.06	47.4	-
UDOP 2023.05	47.4	-
MM1-30B 2024.09	47.3	-
DocKylin 2026.03	46.6	-
DocOwl2 2026.03	46.4	-
T5large+2D+U 2023.06	46.1	-
Cambrian-34B 2024.12	46	-
HVFA 2024.11	45.9	-
mPLUG-Owl-7B + Ours 2024.11	45.9	-
Park et al. 2026.03	45.9	-
MM1-7B 2024.09	45.5	-
LayoutLMv3large 2023.06	45.1	-
MM1-3B 2024.09	44.7	-
CogAgent 2026.03	44.5	-

Showing 100 of 135 rows