Loading...
Loading...
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. Built with scaling-efficient infrastructure for high-throughput training and inference, the model leverages advanced post-training techniques including SFT, DPO, and UPO for optimized performance, while supporting an impressive 131K context length and RLVR alignment for superior cross-modal reasoning and generation capabilities.
Preço/1M
$0.25
305th mais barato
21% abaixo da mediana
Top 45%
Janela de Contexto
30K
323rd maior
Top 92%
Entrada
$0.14
por 1M tokens
Saída
$0.56
por 1M tokens
Combinado
$0.25
por 1M tokens
Mais barato que 55% dos modelos. Preço mediano é $0.31/1M tokens.
Diário
$0.25
Mensal
$7.35
Janela de Contexto
30K
tokens
Maior que 8% dos modelos
Saída Máxima
8K
tokens
27% do contexto
Comparação de Janela de Contexto