DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on...
Rank
#2
Up 122% week over week
Tokens
17.8T
2026-09-22
Requests
294.0M
293,970,473
Tool calls
-
Input / M
$0.10
Output / M
$0.50
Cache / M
$0.01
Context
1.0M
1,048,576 tokens