ERNIE-4.5-VL-424B-A47B-Base
Active Parameters
424B
Context Length
131K
Modality
Multimodal
Architecture
Mixture of Experts (MoE)
License
Apache 2.0
Release Date
30 Jun 2025
Knowledge Cutoff
Jun 2025
API Pricing (per 1M)
Self-hosted only
System Requirements
VRAM requirements for different quantization methods and context sizes
1,024 tokens
Consumer
51x RTX 4090
24GB VRAM
Datacenter
14x NVIDIA A100
80GB VRAM
Apple Silicon
11x Apple M3 Max
128GB VRAM
131,072 tokens
Consumer
53x RTX 4090
24GB VRAM
Datacenter
14x NVIDIA A100
80GB VRAM
Apple Silicon
11x Apple M3 Max
128GB VRAM
Architecture Diagram
Evaluation Benchmarks
No evaluation benchmarks for ERNIE-4.5-VL-424B-A47B-Base available.
Rankings
Overall Rank
-
Coding Rank
-
About ERNIE-4.5-VL-424B-A47B-Base
ERNIE-4.5-VL-424B-Base is Baidu's flagship multimodal Mixture-of-Experts base model activating 47B parameters for cross-modal pretraining. Built with modality-isolated routing, it enables scalable vision-language reasoning over a 131K context window.
Technical Specifications
Attention
Attention Structure
Grouped-Query Attention
Attention Heads
64
Key-Value Heads
8
Attention Head Dimension
128
Position Embedding
Absolute Position Embedding
RoPE Theta
500,000
Sliding Window Attention
No
Sliding Window Size
-
Sliding Window Ratio
-
Linear Attention
-
Linear Attention Ratio
-
Normalization
RMS Normalization
Activation Function
Swish
Dimensions
Hidden Dimension Size
4,096
Number of Layers
54
FFN Intermediate Size (Dense)
28,672
Multi-Token Prediction Heads
-
Tokenizer
Vocabulary Size
103,424
Mixture of Experts
Total Expert Parameters
47.0B
Number of Experts
128
Active Experts
16
Shared Experts
-
FFN Intermediate Size (per Expert)
-
Dense Layers Before MoE
3
About ERNIE 4.5
The Baidu ERNIE 4.5 family consists of ten large-scale multimodal models. They utilize a heterogeneous Mixture-of-Experts (MoE) architecture, which enables parameter sharing across modalities while also employing dedicated parameters for specific modalities, supporting efficient language and multimodal processing.