Alibaba
$2
Input tokens/M
$20
Output tokens/M
-
Context Length
$4
$16
1k
Baidu
128
$1
$10
256
$8
$240
52
$3.9
$15.2
64
$0.8
32
$54
$163
Tencent
Deepseek
$12
$1.6
$0.75
tensorblock
The Mixture of Experts (MoE) model of Tongyi Qianwen version 1.5, with a parameter scale of 2.7B, providing GGUF format files in multiple quantization versions.