Top AI Product

Qwen shipped Qwen3.8-Max-Preview on July 19: a 2.4 trillion parameter model, the team’s first multimodal one past the 1T mark. It reads images, video, and documents. Alibaba says it’s going open-weight — which would make it the largest open multimodal model ever. For now, only the hosted preview exists.

The headline claim: performance “second only to Fable 5.” Bold, because there are no SWE-Bench, GPQA, or Arena scores yet — just internal claims. The timing is transparent: Kimi K3 owns the open-model spotlight, Moonshot is reportedly six months from an IPO, and a free 2.4T alternative hurts exactly that story.

API access

Live on Alibaba Cloud’s Token Plan at 10% of standard pricing during preview, OpenAI-compatible API, also built into Qoder and QoderWork. Self-hosting unlocks once weights actually drop. Alibaba says the biggest gains over Qwen 3.7-Max are in coding — full-stack builds and data analysis.

Why it matters

415 points on HackerNews in one weekend. If the weights land, frontier-minus-one stops being an API-only privilege. If they don’t, “second only to Fable 5” was just a press release.

You Might Also Like

• Kimi k2 6 Beats gpt 5 4 and Claude Opus 4 6 on swe Bench pro
• Alibabas Qwen 3 6 27b Scores 77 2 on swe Bench and Runs on one Consumer gpu
• Alibaba Qwen 3 5 Just Dropped and it Brought 10 Million Milk Teas With it
• Z ais glm 5 1 Tops swe Bench pro at 58 4 Trained on Zero Nvidia Hardware
• Alibaba Qwen Smart Glasses g1 s1 275 ai Glasses With Swappable Batteries and a Qwen api Backend