Our new X account is live! Follow @wizwand_team for updates

Visual Mathematical Reasoning on MathVista

85Accuracy

InternVL3.5-8B-Masters

Updated 3d ago

Evaluation Results

Method
InternVL3.5-8B-Masters 2025.12	85	-	-	-	-	-	-
Masters 2025.12	85	-	-	-	-	-	-
GLM-4.5V 2025.12	84.6	-	-	-	-	-	-
Qwen3-VL-32B 2025.12	83.8	-	-	-	-	-	-
InternVL3-8B-Masters 2025.12	82.3	-	-	-	-	-	-
Masters 2025.12	82.3	-	-	-	-	-	-
Octopus-8B (Ours) 2026.02	82.1	-	-	-	-	-	-
GPT-5 2025.12	81.9	-	-	-	-	-	-
InternVL3.5-38B 2025.12	81.9	-	-	-	-	-	-
MiMo-VL-7B-SFT 2026.02	81.8	-	-	-	-	-	-
Qwen3-VL-8B-Masters 2025.12	81.8	-	-	-	-	-	-
Masters 2025.12	81.8	-	-	-	-	-	-
MiMo-VL-7B-RL 2026.02	81.5	-	-	-	-	-	-
MiMo-VL-8B 2025.12	81.5	-	-	-	-	-	-
Keye-VL-1.5-8B 2025.12	81.2	-	-	-	-	-	-
Qwen3-VL-8B-Instruct + GSPO 2026.02	81	-	-	-	-	-	-
Gemini-2.5-Pro 2025.12	80.9	-	-	-	-	-	-
Qwen3-VL-8B-Instruct + GRPO 2026.02	80.8	-	-	-	-	-	-
GLM-4.1V-9B 2025.12	80.7	-	-	-	-	-	-
Keye-VL-8B 2025.12	80.7	-	-	-	-	-	-
Qwen3-VL-8B-Instruct + DAPO 2026.02	80.3	-	-	-	-	-	-
Qwen3-VL-8B-Thinking 2026.02	79.8	-	-	-	-	-	-
Qwen3-VL-8B-Instruct + SRPO 2026.02	79.8	-	-	-	-	-	-
Qwen3-VL-4B-Masters 2025.12	79.6	-	-	-	-	-	-
Doubao-pro-1.5 2026.01	79.5	-	77.7	88.9	86	82.3	62
Qwen3-VL-8B-Instruct + GRPO 2026.02	79.1	-	-	-	-	-	-
GPT-5-Mini 2025.12	79.1	-	-	-	-	-	-
InternVL3-78B 2025.12	79	-	-	-	-	-	-
Qwen2.5-VL-7B-Masters 2025.12	78.8	-	-	-	-	-	-
InternVL3.5-4B-Masters 2025.12	78.8	-	-	-	-	-	-
Masters 2025.12	78.8	-	-	-	-	-	-
Qwen3-VL-8B-Instruct + GSPO 2026.02	78.5	-	-	-	-	-	-
InternVL3.5-8B 2025.12	78.4	-	-	-	-	-	-
Qwen3-VL-8B-Instruct + SRPO 2026.02	77.9	-	-	-	-	-	-
InternVL3.5-2B-Masters 2025.12	77.5	-	-	-	-	-	-
Qwen3-VL-8B 2025.12	77.2	-	-	-	-	-	-
InternVL3.5-4B 2025.12	77.1	-	-	-	-	-	-
Shuffle-R1-Qwen-7B 2025.08	77	-	-	-	-	-	-
Qwen3-VL-8B-Instruct 2026.02	76.9	-	-	-	-	-	-
COGFLOW 2026.01	76.8	-	70.4	93.1	73.7	86.9	59.3
Qwen2.5-VL-3B-Masters 2025.12	75.6	-	-	-	-	-	-
Claude-4-Sonnet 2025.12	74.6	-	-	-	-	-	-
AdaTooler-V-7B 2025.12	74.5	-	-	-	-	-	-
InternVL3.5-8B-RL 2026.02	74.2	-	-	-	-	-	-
Qwen2.5-VL-72B 2025.12	74.2	-	-	-	-	-	-
NoisyRollout 2025.12	74.16	-	-	-	-	-	-
OpenAI-o1 2026.02	73.9	-	-	-	-	-	-
o1 2025.08	73.9	-	-	-	-	-	-
VL-Rethinker 2026.01	73.7	-	-	-	-	-	-
Qwen3-VL-4B 2025.12	73.7	-	-	-	-	-	-
MiniCPM-o2.6-8B 2025.12	73.3	-	-	-	-	-	-
Active-Zero 2026.02	72.6	-	-	-	-	-	-
ThinkLite-VL-7B 2025.08	72.4	-	-	-	-	-	-
SaEI 2025.12	72.1	-	-	-	-	-	-
VL-Rethinker-7B 2025.08	72	-	-	-	-	-	-
Ovis2-8B 2025.12	71.8	-	-	-	-	-	-
InternVL3.5-2B 2025.12	71.8	-	-	-	-	-	-
Vanilla GRPO 2025.12	71.67	-	-	-	-	-	-
KL-Cov 2025.12	71.6	-	-	-	-	-	-
ThinkLite-VL 2026.01	71.6	-	-	-	-	-	-
InternVL3-8B 2025.12	71.6	-	-	-	-	-	-
InternVL3-8B 2025.12	71.6	-	-	-	-	-	-
InternVL-3-8B 2025.08	71.6	-	-	-	-	-	-
MMR1-Math-7B 2025.08	71.5	-	-	-	-	-	-
SophiaVL-R1 2026.01	71.3	-	-	-	-	73.4	-
SophiaVL-R1-7B 2025.12	71.3	-	-	-	-	-	-
Gemini-2.0 pro 2025.08	71.3	-	-	-	-	-	-
NoisyRollout-7B-K12 2025.08	70.9	-	-	-	-	-	-
Vision-R1-7B 2025.08	70.8	-	-	-	-	-	-
MM-Eureka-Qwen-7B 2025.08	70.6	-	-	-	-	-	-
EvoLMM 2026.01	70.52	-	-	-	-	-	-
Qwen3-VL-2B-Masters 2025.12	70.4	-	-	-	-	-	-
Gemini-2.0-Flash 2025.12	70.4	-	-	-	-	-	-
GPT-4.1 2025.12	70.4	-	-	-	-	-	-
Shuffle-R1-Qwen-3B 2025.08	70.4	-	-	-	-	-	-
VisionZero-CLEVR 2026.02	70.2	-	-	-	-	-	-
DeepEyes 2025.12	70.1	-	-	-	-	-	-
Thymes 2025.12	70	-	-	-	-	-	-
VLAA-Thinker-7B 2025.08	69.9	-	-	-	-	-	-
EvoLMM 2026.02	69.8	-	-	-	-	-	-
IREASONER 2026.01	69.74	-	-	-	-	-	-
OpenVLThinker-7B 2025.08	69.7	-	-	-	-	-	-
LLaVA-OneVision-1.5-8B 2025.12	69.6	-	-	-	-	-	-
Ovis2-4B 2025.12	69.6	-	-	-	-	-	-
Base Model 2026.02	69.4	-	-	-	-	-	-
VisPlay 2026.02	69.4	-	-	-	-	-	-
Qwen2.5-VL-7B w/ Discrete Reward + Step-level Majority 2026.01	69.31	-	-	-	-	-	-
VisionZero-RealWorld 2026.02	69.2	-	-	-	-	-	-
Qwen2.5-VL-7B w/ Discrete Reward 2026.01	68.88	-	-	-	-	-	-
Qwen2.5-VL-7B (Baseline) 2026.01	68.47	-	-	-	-	-	-
Vision-Zero 2026.01	68.43	-	-	-	-	-	-
VisionZero-Chart 2026.02	68.2	-	-	-	-	-	-
Qwen2.5-VL-7B-Instruct 2025.12	68.2	-	-	-	-	-	-
LLaVA-OneVision-1.5-4B 2025.12	67.9	-	-	-	-	-	-
Qwen2.5-VL-7B 2025.12	67.8	-	-	-	-	-	-
Claude-3.5-Sonnet 2026.01	67.7	-	-	-	-	-	-
Claude-3.5-Sonnet 2025.12	67.7	-	-	-	-	-	-
LLaVA-OneVision-72B 2025.12	67.5	-	-	-	-	-	-
Qwen2.5-VL-7B 2025.08	67.4	-	-	-	-	-	-
Qwen2.5-VL-7B-Instruct 2025.12	67.14	-	-	-	-	-	-

Showing 100 of 194 rows