Alibaba launches Qwen4-Max: native vision, 1M context, 60% cheaper per task
Kyrannio · x · 2026-11-16
Alibaba's Qwen team released Qwen4-Max, a new flagship built on a fresh MoE architecture with native vision and a 1M-token context window, positioned as an "AI teammate" for work, creativity, and discovery.
- Cost: up to 50% fewer tokens than Qwen3.8-Max per task, with average cost per task down over 60% in internal evals
- Capabilities: stronger professional tool use, long-horizon execution, and deliverable quality across 3D modeling, game dev, manufacturing, finance, law, healthcare, and scientific research
- Pricing: $1.5/M input tokens, $4.5/M output tokens, $0.075/M implicit caching
The official benchmark charts show leads on agentic coding suites like DeepSWE, CodeArena, and Terminal-Bench; reposter kimmonismus cautions that leaderboard numbers and real-world use remain two different things.
More from Models
- Qwen4-Max lands #6 in Image-to-WebDev Arena at 1674 pts, up 35 over predecessor — Alibaba_Qwen · 2026-11-16
- Users can't access any Gemini model in Google's Antigravity, forced to use Claude — rez0__ · 2026-11-16
- Qwen4 Max per-eval breakdown: strong on knowledge work, lagging in coding — ArtificialAnlys · 2026-11-16
- Qwen4 Max scores 43 on AA Intelligence Index, 67% cheaper per task than predecessor — ArtificialAnlys · 2026-11-16
- Qwen4-Max debuts at #14 on Agent Arena, making Alibaba the #5 lab overall — Alibaba_Qwen · 2026-11-16
- New MoE base claims win over Qwen3.8-Max with 50% fewer tokens, met with skepticism — teortaxesTex · 2026-11-16