ERNIE 4.5 VL 424B A47B

ERNIE 4.5 VL 424B A47B is Baidu's flagship multimodal vision-language model, using a Mixture-of-Experts architecture with 424B total parameters and 47B activated per token. It accepts text and image inputs, supports a 131,072-token context window, and is instruction-tuned via SFT and reinforcement learning for visual reasoning, document understanding, and visual question answering. Source: supabase.