← All models
B

Baidu: ERNIE 4.5 VL 28B A3B

baidu/ernie-4.5-vl-28b-a3b

VisionTool useReasoningStreaming

A powerful multimodal Mixture-of-Experts chat model featuring 28B total parameters with 3B activated per token, delivering exceptional text and vision understanding through its innovative heterogeneous MoE structure with modality-isolated routing....

Pricing

Input

$0.14 / 1M

Output

$0.56 / 1M

Specs

Context

131,072 tokens

Input

text, image

Output

text

Knowledge cutoff: 2025-03-31

Released: 2025-08

Supported parameters

frequency_penaltyinclude_reasoningmax_tokenspresence_penaltyreasoningrepetition_penaltyseedstoptemperaturetool_choicetoolstop_ktop_p

Open weights · HuggingFace

197,916 downloads/mo
103 likes
apache-2.0 image-text-to-text
View on HuggingFace →