What changed
2026-08-03 12:00 UTC → 2026-08-03 15:11 UTC ·
added
removed
Alibaba launches 2.4T Qwen3.8‑Max model expands Qwen3.8 AI offerings
After June’s zero‑configuration deployment guides and Following the July announcement unveiling of the 2.4‑trillion‑parameter Qwen 3.8 model, which Alibaba expanded its AI portfolio in early August with said ranks second only to Anthropic’s Claude Fable 5, the public beta company has rolled out a series of QwenWork, products and pricing options around the model family. On 19 July Alibaba Cloud introduced a SaaS office platform that embeds Token Plan personal edition for the Qwen 3.8 and offers free 3.8‑Max‑Preview model, offering Lite (39 CNY), Standard (139 CNY) and paid plans for individuals Pro (499 CNY) monthly tiers with credit limits, concurrent‑agent caps and enterprises. On 3 daytime/night‑time discounts. The same day the firm announced the Qwen 3.8 model itself, promising an open‑weight release in the near future. In early August 2026 Alibaba launched the company unveiled Qwen3.8‑Max, Qwen 3.8‑Max flagship, a multimodal successor mixture‑of‑experts system that processes activates about 95 billion parameters per request, handles text, images, video and up to one million tokens per request. Built on a mixture‑of‑experts architecture, it activates roughly 95 billion parameters per query, lowering latency and compute cost. In internal testing the model completed an autonomous software‑engineering project in 16 days tokens, and achieved benchmark scores comparable to Anthropic’s Claude Fable 5, ranking second worldwide on the Arena AI leaderboard for image‑analysis and becoming the highest‑ranking Chinese model for text tasks. Alibaba announced that the model’s weights will be publicly 5. Weights were made downloadable the week of 10 August via Alibaba Cloud’s Model Studio, and it set token usage pricing at US$2 per million input tokens and US$6 per million output tokens. The launch the announcement lifted Alibaba’s Hong Kong‑listed Kong shares by 5‑7 % up to 7 percent. The company also opened public beta testing of QwenWork, an enterprise AI office suite built on Qwen 3.8, with free and the firm paid plans to extend its shift ranging from on‑premise guides ¥78 to cloud‑based, token‑priced subscriptions ¥198 per month and a full SaaS AI suite built around the Qwen 3.8 family. flagship enterprise option on request.