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

Mathematical Reasoning on MathVista (Accuracy)

89.2Accuracy

Gemini 3-Pro

Updated 12d ago

Evaluation Results

Method
Gemini 3-Pro 2026.02	89.2	-	-
Gemini3-pro 2026.05	88.5	-	-
Qwen3-VL 2026.02	86.8	-	-
Doubao-Seed-1.6 2026.05	85.9	-	-
Qwen3-VL-235B-Thinking 2026.05	85.9	-	-
TwC (Ours) - Img & Txt 2026.02	85.8	-	-
ERNIE 5.0 2026.02	84.8	-	-
GLM-4.5V 2026.05	84.6	-	-
ERNIE 5.0-Base 2026.02	84.4	-	-
Qwen3-VL-32B 2025.06	83.8	-	-
Gemini 2.5-Pro 2026.02	82.7	-	-
GPT-5 2026.02	82.1	-	-
InternVL3.5-38B 2025.06	81.9	-	-
GPT-5-Thinking* 2025.09	81.9	-	-
GPT-5-20250807 2026.05	81.9	-	-
GenRecal (InternVL3.5-8B) 2025.06	81.5	-	-
GenRecal (InternVL3.5-8B) 2025.06	81.3	-	-
Gemini-2.5-Pro* 2025.09	80.9	-	-
GenRecal (Qwen3-VL-8B) 2025.06	80.3	-	-
GenRecal (Qwen3-VL-8B) 2025.06	80.1	-	-
Kimi-vl-A3B-thinking 2026.05	80.1	-	-
InternVL3.5-8B 2025.06	78.4	-	-
Qwen3-VL-8B-Instruct + LLM-as-a-Judge 2026.05	77.6	-	-
DEITA 2026.05	77.5	-	-
Penguin-VL 2026.03	77.4	-	-
OST 2026.05	77.4	-	-
Qwen3-VL-8B 2025.06	77.2	-	-
Qwen3-VL 2026.03	77.2	-	-
Qwen3-VL-8B-Instruct 2026.05	77.2	-	-
HEED 2026.05	77.2	-	-
VL-Cogito-7B + LEAD 2026.03	76.3	-	-
Qwen3-VL-8B-Instruct + Random 2026.05	76.1	-	-
Teacher 2026.05	76	-	-
OST 2026.05	75.8	-	-
VL-Rethinker-7B + LEAD 2026.03	75.6	-	-
VAPO-Thinker-7B 2025.09	75.6	-	-
RSA 2026.05	75.4	-	-
Qwen3-VL-4B-Instruct + LLM-as-a-Judge 2026.05	75.3	-	-
ThinkLite-VL 2026.04	75.1	-	-
DEITA 2026.05	75.1	-	-
TwC (Ours) - Only Image 2026.02	75	-	-
InternVL2.5-8B-GenRecal 2025.06	74.9	-	-
Vision-R1-7B + LEAD 2026.03	74.9	-	-
VL-Rethinker-7B 2026.03	74.9	-	-
VL-Rethinker 2026.04	74.9	-	-
V-STAR 2026.04	74.9	-	-
Qwen3-VL-4B-Instruct 2026.05	74.9	-	-
VL-Cogito-7B 2026.03	74.8	-	-
VL-Cogito 2026.04	74.8	-	-
Qwen2.5-VL-32B 2026.02	74.7	-	-
GeoFocus 2026.02	74.3	-	-
InternVL-3.5 2026.03	74.2	-	-
Qwen3-VL-4B-Instruct + Random 2026.05	74.2	-	-
RuCL 2026.02	74.1	-	-
HSA 2026.05	74.1	-	-
MM-Eureka-Qwen-32B 2025.11	74	-	-
OpenAI-o1 2026.05	73.9	-	-
Vision-R1-7B 2026.01	73.5	-	-
Vision-R1-7B 2026.03	73.5	-	-
Vision-R1-7B 2025.09	73.5	-	-
Vision-R1 2026.04	73.5	-	-
MM-Eureka-Qwen-7B 2025.11	73.5	-	-
Vision-R1-7B 2026.05	73.5	-	-
Qwen2.5-VL-7B-Instruct + Faithful-MR1 2026.05	73.5	-	-
KD 2026.05	73.4	-	-
ThinkLite-VL-7B 2026.02	73.3	-	-
GazeVLM (Ours) 2026.05	73.3	-	-
VL-Rethinker-7B 2026.02	73.27	-	-
Qwen3-VL-4B-Instruct + Full SFT 2026.05	73.2	-	-
Perception-R1-7B 2026.05	73.2	-	-
MM-Eureka-7B 2026.02	73	-	-
MM-Eureka-Qwen-7B 2025.11	73	-	-
Perception-R1-7B 2026.02	72.8	-	-
VL-Rethinker 2026.04	72.8	-	-
AutoTool 2026.05	72.8	-	-
Claude-Sonnet 4.5 2026.02	72.5	-	-
VL-Rethinker-7B 2025.11	72.4	-	-
InternVL2.5-78B 2025.06	72.3	-	-
OpenVLThinker 2026.04	72.3	-	-
GRPO 2026.02	72.1	-	-
MM-Eureka-Qwen-32B 2025.11	72.1	-	-
VL-Rethinker-7B 2025.11	72	-	-
InternVL2.5-8B-GenRecal 2025.06	71.9	-	-
DeepEyesV2 2026.03	71.9	-	-
MMR1-Math-v0-7B 2025.11	71.9	-	-
InternVL2.5-38B 2026.02	71.84	-	-
InternVL3.5-2B 2025.12	71.8	-	-
Qwen3-VL-4B 2026.05	71.7	-	-
InternVL3 2026.03	71.6	-	-
Semantic-back 2026.03	71.6	-	-
InternVL3.5-8B 2026.04	71.6	-	-
DeepEyes 2026.05	71.6	-	-
Gemini-3-Pro 2026.02	71.5	-	-
InternVL2.5-8B-GenRecal 2025.06	71.4	-	-
OpenVLThinker-7B 2026.02	71.38	-	-
VRE 2026.03	71.2	-	-
Video-R1-7B 2026.01	71	-	-
MMR1-Math-v0-7B 2025.11	71	-	-
Qwen2.5-VL-7B-Instruct + GRPO 2026.05	70.9	-	-
Vision-R1-7B 2026.02	70.63	-	-

Showing 100 of 382 rows