Alibaba's Qwen3.8-Max Leaks: 2.4T Parameters and 1M Context
Aiden_Tech_Ai · x · 2026-08-05
Specs for Alibaba's rumored new flagship model, Qwen3.8-Max, have surfaced. The model reportedly features 2.4 trillion total parameters with a Sparse MoE architecture activating only 95B parameters. It supports a 1 million-token context window and native multimodal reasoning across vision and text.
Designed for the agentic AI era, the model is said to handle autonomous coding, long-horizon tasks, massive document analysis (200+ pages), and complex research workflows. It is reportedly accessible globally via the Alibaba Cloud Model Studio API.
More from Models
- OvisOCR2 Tested: High Performance in Small Size Thanks to Data and Training Recipe — vanstriendaniel · 2026-08-05
- Developer accidentally tweaks Gemma layer, changes its favorite color — dejanseo · 2026-08-05
- Developer: DeepSeek Flash is genuinely good, failures are just intellectual limits — yacineMTB · 2026-08-05
- Alt sources for MiniMax H3 quants and LoRAs — reeight · 2026-08-05
- Ridiculous Experience: User Kicked from Fable to Sonnet to Haiku — basedjensen · 2026-08-05
- Xiaomi Hires DeepSeek Core Researcher Luo Fuli with Tens of Millions Salary to Lead AI — thisdudelikesAI · 2026-08-05