This is the batch-processing variant of Qwen3.8-2.4T-A95B, Alibaba's open-weight sparse MoE model (2.4T total parameters, 95B active, 1M-token context) released under Alibaba's custom Qwen3.8-Max license as the open-weight counterpart to the closed flagship Qwen3.8 Max. The batch endpoint queues requests for cheaper asynchronous processing, suited to large-scale offline coding, research, or agentic workloads where latency does not matter.
Taux de coût
1x
Contexte
1M
Date de sortie
12 août 2026
Entrée
Text
Sortie
Text
Prise en charge
RaisonnementAppel d'outilsSorties structurées