What happened
Qwen says Qwen4-Max is live on You.com and invites users to try it. The quoted You.com post says the model is available on Inference Router from Day 0 and describes it as built for coding, office work, and long-horizon agent runs, with a 1M-token context and native text, image, and video input.
Qwen4-Max is live on @youdotcom. 🚀 Try it now!
View the quoted post on XQwen 4-Max is live on Inference Router, from Day 0.
The first model in @Alibaba_Qwen's Qwen4 series, built for coding, office work, and long-horizon agent runs. 1M-token context, native text, image, and video input.
Try it now: you.com/product/inference-rou…
The story so far
Alibaba has launched Qwen4-Max, a 2.4T-parameter MoE foundation model featuring native vision capabilities and a 1M-token context window. The model is now available via the Qianwen AI platform API and integrated into various services, including Token Plan, Qwen Studio, QwenWork, Qoder, Qwen APP, OpenRouter, OpenCode Go, Vercel AI Gateway, Novita Labs, You.com, LebrelAI, and Runware. Alibaba plans to release open weights for smaller models in the Qwen4 series over time. Qwen states that Qwen4-Max outperforms Qwen3.8-Max on the Work360 benchmark with a score of 52.69, a 10% increase, while reporting significant reductions in token usage and average task costs. On Terminal-Bench 4.0, QbitAI reports a score of 48.5 against 38.9 for the previous model, with average task costs declining by 58%. API pricing is set at $1.5 per million input tokens, $4.5 per million output tokens, and $0.075 per million tokens for implicit caching. Vercel notes its implementation supports up to 384K output tokens and is priced 25% lower than Qwen3.8-Max, while QbitAI cites cache-hit pricing as low as 0.5 yuan per million tokens on the Qianwen AI platform.
How the story developed · 3
View full event- 11/17 12:00 (UTC+8)
Qwen4-Max Goes Live on You.com This report
- 11/16 22:39 (UTC+8)Alibaba Releases Qwen4-Max with 2.4T Parameters and 1M-Token ContextQbitAI
- 11/16 10:38 (UTC+8)Qwen4-Max launches with native vision and a 1M-token context windowQwen · X
