DocVQA

Overview

Type: percentage

Recorded Results

Average Score

93.36%

Score Range

85.60% - 96.40%

Leading Model (lowest score)

85.60% - Grok 1.5

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
SpaceXAI	Grok 1.5	28 Mar 2024	85.60%	-	Yes	Source
SpaceXAI	Grok 1.5V	12 Apr 2024	85.60%	-	Yes	Source
Meta	Llama 3.2 11B Vision Instruct	-	88.40%	inferred high-confidence family alias from llama-3.2-11b-instruct (score=0.5873; benches=11)	Yes	Source
Meta	Llama 3.2 1B Instruct	25 Sept 2024	88.40%	inferred family alias from llama-3.2-11b-instruct (score=0.4200; benches=11)	Yes	Source
DeepSeek	DeepSeek VL2 Tiny	13 Dec 2024	88.90%	-	Yes	Source
Meta	Llama 3.2 90B Vision Instruct	-	90.10%	inferred high-confidence family alias from llama-3.2-90b-instruct (score=0.5873; benches=13)	Yes	Source
Mistral	Pixtral 12B	17 Sept 2024	90.70%	inferred version-family alias from pixtral-12b-2409	Yes	Source
DeepSeek	DeepSeek VL2 Small	13 Dec 2024	92.30%	-	Yes	Source
OpenAI	GPT 4o (2024-08-06)	06 Aug 2024	92.80%	-	Yes	Source
OpenAI	GPT 4o Audio (2025-06-03)	03 Jun 2025	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Audio (2024-10-01)	01 Oct 2024	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Audio (2024-12-17)	17 Dec 2024	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Realtime Preview (2025-06-03)	03 Jun 2025	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Realtime Preview (2024-10-01)	01 Oct 2024	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Search Preview	11 Mar 2025	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Transcribe	20 Mar 2025	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
OpenAI	GPT 4o Transcribe Diarize	15 Oct 2025	92.80%	inferred modality/version alias from gpt-4o-2024-08-06	Yes	Source
Microsoft	Phi 4 multimodal instruct	01 Feb 2025	93.20%	-	Yes	Source
SpaceXAI	Grok 2 Mini	13 Aug 2024	93.20%	-	Yes	Source
DeepSeek	DeepSeek VL2	13 Dec 2024	93.30%	-	Yes	Source
Mistral	Pixtral Large	18 Nov 2024	93.30%	-	Yes	Source
SpaceXAI	Grok 2	13 Aug 2024	93.60%	-	Yes	Source
Meta	Llama 4 Maverick	05 Apr 2025	94.40%	-	Yes	Source
Meta	Llama 4 Scout	05 Apr 2025	94.40%	-	Yes	Source
Qwen	Qwen 2.5 Coder 32B Instruct	-	94.80%	inferred high-confidence family alias from qwen2.5-vl-32b (score=0.4641; benches=28)	Yes	Source
Qwen	Qwen 2.5 VL 32B Instruct	24 Mar 2025	94.80%	-	Yes	Source
Qwen	Qwen 2.5 VL 3B Instruct	-	94.80%	inferred high-confidence family alias from qwen2.5-vl-32b (score=0.4914; benches=28)	Yes	Source
Mistral	Mistral Small 3.2	20 Jun 2025	94.86%	-	Yes	Source
Qwen	Qwen 2.5 Coder 3B	-	95.20%	inferred family alias from qwen2.5-omni-7b (score=0.3000; benches=45)	Yes	Source
Qwen	Qwen 2.5 Coder 7B	-	95.20%	inferred high-confidence family alias from qwen2.5-omni-7b (score=0.4700; benches=45)	Yes	Source
Qwen	Qwen 2.5 Math 7B	-	95.20%	inferred high-confidence family alias from qwen2.5-omni-7b (score=0.4767; benches=45)	Yes	Source
Qwen	Qwen 2.5 Math 7B PRM800K	-	95.20%	inferred family alias from qwen2.5-omni-7b (score=0.3696; benches=45)	Yes	Source
Qwen	Qwen 2.5 Math PRM 7B	-	95.20%	inferred family alias from qwen2.5-omni-7b (score=0.4092; benches=45)	Yes	Source
Qwen	Qwen 2.5 Omni 3B	-	95.20%	inferred high-confidence family alias from qwen2.5-omni-7b (score=0.4933; benches=45)	Yes	Source
Qwen	Qwen 2.5 Omni 7B	-	95.20%	-	Yes	Source
Qwen	Qwen 2.5 0.5B	-	95.70%	inferred family alias from qwen2.5-vl-7b (score=0.3131; benches=32)	Yes	Source
Qwen	Qwen 2.5 1.5B	-	95.70%	inferred family alias from qwen2.5-vl-7b (score=0.3131; benches=32)	Yes	Source
Qwen	Qwen 2.5 3B	-	95.70%	inferred family alias from qwen2.5-vl-7b (score=0.3531; benches=32)	Yes	Source
Qwen	Qwen 2.5 VL 7B Instruct	-	95.70%	-	Yes	Source
Qwen	Qwen 2.5 Math 72B	-	96.40%	inferred high-confidence family alias from qwen2.5-vl-72b (score=0.4700; benches=30)	Yes	Source
Qwen	Qwen 2.5 Math PRM 72B	-	96.40%	inferred family alias from qwen2.5-vl-72b (score=0.4050; benches=30)	Yes	Source
Qwen	Qwen 2.5 Math RM 72B	-	96.40%	inferred family alias from qwen2.5-vl-72b (score=0.4092; benches=30)	Yes	Source
Qwen	Qwen 2.5 VL 72B Instruct	26 Jan 2025	96.40%	-	Yes	Source