Summary
According to a post on X by @Astrodevil_, Qwen4-Max has come out of preview. The post describes a MoE model with 2.4T total and 71B active parameters, native vision and a 1M-token context. Citing Artificial Analysis runs at xhigh effort, the post reports an Intelligence Index of 43, up from 34 for Qwen3.8-Max. It also reports about 105k output tokens per task (down from about 171k), $3.27 per task (down from $10.02), and a 25% lower per-token price.
AI-assisted summary based on the listed source.
What happened
@Astrodevil_ summarized Qwen4-Max coming out of preview: a MoE model with 2.4T total and 71B active parameters, native vision and a 1M-token context. Citing Artificial Analysis runs at xhigh effort, he reports an Intelligence Index of 43 (up from 34 for Qwen3.8-Max), about 105k output tokens per task (down from...
Why it matters
For developers and businesses comparing models, the reported figures point to a lower cost per task alongside a higher benchmark score. These numbers come from a single social post summarizing one evaluation setup, so teams should check the original Artificial Analysis results and their methodology. The source snippet is cut off before it states the model's availability terms, which remains unclear.
Signal Intelligence
Signal Strength 95%
Technical label WATCH
Public Interest 43
Category OPEN SOURCE
Reader Depth TECHNICAL
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 51
Practical Impact Score 28
Novelty Interest Score 70
Consequence Score 34
Curiosity Score 0
Shareability Score 57
Why this is here
VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Scobleizer News Pulse (X).