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

Counting on TallyQA

82.2Accuracy

GLM-4.6V

Updated 1d ago

Evaluation Results

Method	Links
GLM-4.6V 2026.05		82.2
MetaForge 2026.06		81.17
Molmo 2026.01		80.8
Gemini-2.5-Flash 2026.05		80.6
MetaForge 2026.06		80.52
Qwen2.5-VL 2026.01		80
MAESTRO 2026.05		79.8
LLaVA-OneVision 2026.01		79.7
PaliGemma 2026.01		79.4
VTOOL-R1 2026.05		79.4
GPT-5 2026.05		79.2
Gemini-2.5-Pro 2026.05		79
Qwen3-VL-32B 2026.05		78.6
Kimi-K2.5 2026.05		78.4
GPT-4o 2026.05		77.8
INTERNVL3-8B + PGT 2026.05		77.4
INTERNVL3-8B 2026.05		77.2
Direct Answering 2026.05		74.8
Untrained Model 2026.05		74.4
THINKLITE-VL 2026.05		73.9
QWEN2.5-VL-7B + PGT 2026.05		73.5
MathCoder-VL 2026.05		73.4
LLaVA-NeXT 2026.01		73.3
Thyme 2026.05		73.2
DeepEyes 2026.05		73
GPT-5.4 2026.06		72.95
SmolVLM2 2026.01		72.9
VTS-V 2026.05		72.8
Chain-of-Focus 2026.05		72.6
QWEN2.5-VL-7B 2026.05		72.4
QWEN2.5-VL-7B + SPECIALIZED MIX 2026.05		72.4
PixelReasoner 2026.05		71.8
LLaVA-AIMv2 2026.01		71
Visual-ARFT 2026.05		70.8
LLaVA v1.5 2026.01		70.7
DeepEyes-v2 2026.05		70.6
IMAGE JIGSAW 2026.05		70.4
VisionReasoner 2026.05		70.2
LLaVA-AIMv2-2D-RoPE 2026.01		69
LLaVA-SigLIP2 2026.01		68.4
QWEN2.5-VL-3B + PGT 2026.05		67.5
LLaVA-SigLIP 2026.01		67.2
QWEN2.5-VL-3B 2026.05		66.4
LLaVA-2D-RoPE 2026.01		65.4
VIGORL-3B 2026.05		64.8
LLaVA-SigLIP2-2D-RoPE 2026.01		64.6
LLAVA-NEXT-LLAMA3-8B + PGT 2026.05		63.9
QWEN2.5-VL-3B + SPECIALIZED MIX 2026.05		63.4
LLaVA-SigLIP-2D-RoPE 2026.01		61.6
LLAVA-NEXT-7B + PGT 2026.05		60.1
LLAVA-NEXT-LLAMA3-8B 2026.05		59.2
GPT-5-mini 2025.12		55.4
GPT-5-mini 2025.12		55.2
Gemma3-4b-it 2026.01		52.5
VALOR 2025.12		51
VALOR 2025.12		51
Claude-3.5-Haiku 2025.12		50.1
Qwen3-8B 2025.12		50.1
GPT-4o 2025.12		49.9
VALOR-RL 2025.12		49.5
Gemini-2.5-Flash 2025.12		48.9
Gemma-3-12B 2025.12		48.9
o4-mini 2025.12		47.9
Llama-3.2-11B 2025.12		47.7
LLAVA-NEXT-7B 2026.05		45.2
Gemini-2.0-Flash 2025.12		43.6
SPATIAL-LADDER-3B 2026.05		43.1