ModelWatch
← Catalog

Qwen3.8 2.4T A95B

Alibaba (Qwen)qwenlm.github.io

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. It is...

long-contextopen-weights
Input price
$2 / 1M tokens
Output price
$6 / 1M tokens
Context window
1.05M tokens
Max output
131K tokens
Output speed
42 tokens/s
Time to first token
5.07s
Inputs
text
Open weights
Yes
OpenRouter id
qwen/qwen3.8-2.4t-a95b
Listed
Aug 12, 2026

Benchmarks

Intelligence Index
n/a
Engineering Index
n/a

Speed and benchmarks for “Qwen3.8 2.4T A95B” from Artificial Analysis. Prices, context window and listing date are live from OpenRouter.

Compare with others