Unverified50% confidenceFactExact time
The study evaluated 13 VLMs spanning five major mainstream families—GPT, Gemini, Gemma, Mistral, and Qwen—with parameter counts ranging from 3B to 32B.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/5/2026
First Seen
Valid until: 1/3/2027
Sources
Related Entities
Related Claims
UnverifiedThe dataset covers four open-weight models — Qwen2.5-7B, Mistral-7B-v0.3, Phi-3.5-mini, and Qwen2.5-1.5B — with 30 trajectories each.71% similarVerifiedQwen3-4B is already on par with DeepSeek V3 or GPT-4o on multiple reasoning evaluation metrics.69% similarUnverifiedThe open-source model MiniMax M2.6 Pro scored 46 points, while GLM 5.3 and Qwen 3.8 Max scored approximately 45 points, both close to GPT-6 Sol's score of 48 points.69% similarUnverifiedSmall VLMs such as LLaVA-7B and Qwen-VL-Chat typically have parameter counts in the billions, while object detection models typically range from tens of millions to hundreds of millions of parameters.67% similarVerifiedOn mainstream benchmarks such as MMLU and HumanEval, the score gaps among Claude 3.5, GPT-4o, Gemini 1.5 Pro, and Grok-2 have been compressed to within the range of statistical error.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/977919API
curl https://kongchang.com/api/v1/knowledge/claims/977919MCP
get_claim(id=977919)