List your company

Document Visual Question Answering on SlideVQA

0.849AccuracySlideAgent
0.118920.308460.4980.68754Oct 30, 2025Dec 1, 2025Jan 3, 2026Feb 4, 2026Mar 9, 2026Apr 10, 2026May 13, 2026

Evaluation results

MethodLinks
SlideAgentMethod Type=Multimodal RAG (Type 2) and Agentic Methods (Type 3), Base Model=GPT-4o2025.100.8490.9050.804
Gemini-2.5-FlashMethod Type=Multimodal LLMs (Type 1)2025.100.8380.9180.783
ViDoRAGMethod Type=Multimodal RAG (Type 2) and Agentic Methods (Type 3), Base Model=GPT-4o2025.100.8110.8810.764
VDocRAGMethod Type=Multimodal RAG (Type 2) and Agentic Methods (Type 3), Base Model=GPT-4o2025.100.80.8780.75
COLPALIMethod Type=Multimodal RAG (Type 2) and Agentic Methods (Type 3), Base Model=GPT-4o2025.100.7880.8340.737
Claude-4.1-OpusMethod Type=Multimodal LLMs (Type 1)2025.100.7840.8230.743
VisRAGMethod Type=Multimodal RAG (Type 2) and Agentic Methods (Type 3), Base Model=GPT-4o2025.100.7820.8540.731
GPT-4oMethod Type=Multimodal LLMs (Type 1)2025.100.770.840.721
Gemini-2.0-FlashMethod Type=Multimodal LLMs (Type 1)2025.100.750.7980.713
Gemini-2.5-Flash-LiteMethod Type=Multimodal LLMs (Type 1)2025.100.7120.870.608
Claude-3.7-SonnetSize=-, Paradigm=End-to-End2025.120.6290.763—
Claude-3.5-HaikuMethod Type=Multimodal LLMs (Type 1)2025.100.6250.5460.683
VimRAGBackbone=Qwen3-VL-8B-Instruct2026.020.624——
GPT-4.1Size=-, Paradigm=End-to-End2025.120.6150.747—
CogDocSize=7B, Paradigm=Unified Thinking2025.120.5830.679—
MoCASize=7B2026.050.583——
Mem1Backbone=Qwen3-VL-8B-Instruct2026.020.557——
GPT-4oSize=-, Paradigm=End-to-End2025.120.5310.658—
GPT-4oSize=-2026.050.531——
VimRAGBackbone=Qwen3-VL-4B-Instruct2026.020.528——
FRAGSize=8B, Paradigm=RAG2025.120.5240.639—
InternVL3-8BSize=8B, Paradigm=End-to-End2025.120.5160.644—
VisRAGSize=8B, Paradigm=RAG2025.120.5090.524—
ReActBackbone=Qwen3-VL-8B-Instruct2026.020.5——
Mem1Backbone=Qwen3-VL-4B-Instruct2026.020.498——
M3DocRAGSize=10B, Paradigm=RAG2025.120.4880.579—
mPLUG-Owl3Size=7B2026.050.488——
Vanilla RAGBackbone=Qwen3-VL-8B-Instruct2026.020.485——
GPT-4o miniSize=-, Paradigm=End-to-End2025.120.4810.607—
GPT-4o-miniSize=-2026.050.481——
MiMo-VL-7B-SFTSize=7B, Paradigm=End-to-End2025.120.480.619—
MemAgentBackbone=Qwen3-VL-8B-Instruct2026.020.453——
ReActBackbone=Qwen3-VL-4B-Instruct2026.020.449——
Vanilla RAGBackbone=Qwen3-VL-4B-Instruct2026.020.448——
MemAgentBackbone=Qwen3-VL-4B-Instruct2026.020.431——
Pixel ReasonerSize=7B2026.050.421——
VDocRAGSize=8B, Paradigm=RAG2025.120.420.44—
Qwen2.5-VL-7B-InstructSize=7B, Paradigm=End-to-End2025.120.3890.498—
Qwen2.5-VL-InstructSize=7B2026.050.389——
Qwen2.5-VL-72B-InstructSize=72B, Paradigm=End-to-End2025.120.3880.544—
Qwen2.5-VL-InstructSize=72B2026.050.388——
VL-RethinkerSize=7B2026.050.385——
DocopilotSize=8B, Paradigm=End-to-End2025.120.3570.431—
DocopilotSize=8B2026.050.357——
VideoRAGBackbone=Qwen3-VL-8B-Instruct2026.020.355——
Llava-OVSize=7B2026.050.35——
VideoRAGBackbone=Qwen3-VL-4B-Instruct2026.020.34——
DeepEyesSize=7B2026.050.33——
SV-RAGSize=8B, Paradigm=RAG2025.120.320.343—
mPLUG-DocOwl2Size=8B, Paradigm=End-to-End2025.120.2540.304—
R1-VLSize=7B2026.050.221——
UniversalRAGBackbone=Qwen3-VL-8B-Instruct2026.020.172——
UniversalRAGBackbone=Qwen3-VL-4B-Instruct2026.020.147——
Claude-3.7-SonnetBackbone=-, Param=-, Paradigm=E2E2026.04—0.763—
CogDocBackbone=Qwen2.5-VL, Param=7B, Paradigm=Agent2026.04—0.679—
Doc-V* (GRPO)Backbone=Qwen2.5-VL, Param=7B, Paradigm=Agent2026.04—0.772—
Doc-V* (SFT)Backbone=Qwen2.5-VL, Param=7B, Paradigm=Agent2026.04—0.738—
DocopilotBackbone=InternVL2, Param=8B, Paradigm=E2E2026.04—0.431—
GPT-4.1Backbone=-, Param=-, Paradigm=E2E2026.04—0.747—
GPT-4oBackbone=-, Param=-, Paradigm=E2E2026.04—0.658—
GPT-4o miniBackbone=-, Param=-, Paradigm=E2E2026.04—0.607—
InternVL3Backbone=InternViT / Qwen2.5, Param=8B, Paradigm=E2E2026.04—0.644—
M3DocRAGBackbone=Qwen2-VL, Param=7B, Paradigm=RAG2026.04—0.557—
mPLUG-DocOwl2Backbone=ViT / LLaMa, Param=8B, Paradigm=E2E2026.04—0.278—
Qwen2.5-VL (Baseline)Backbone=Qwen2.5-VL, Param=7B, Paradigm=E2E2026.04—0.552—
Qwen2.5-VL (RAG Top-5)Backbone=Qwen2.5-VL, Param=7B, Paradigm=RAG2026.04—0.629—
SV-RAGBackbone=InternVL2, Param=4B, Paradigm=RAG2026.04—0.343—
VDocRAGBackbone=Phi3-Vision, Param=4B, Paradigm=RAG2026.04—0.42—
VisRAGBackbone=MiniCPM-V 2.6, Param=8B, Paradigm=RAG2026.04—0.524—