Alibaba launches Qwen4-Max: 2.4T MoE, 71B activations, 1M context
Automated summary · Based on collected summaries; not individually verified
Prepared by HotAI
What happened
Alibaba's Qwen team launches Qwen4 series' first model, Qwen4-Max: a proprietary sparse MoE architecture with 2.4T backbone parameters, 71B activations per token, plus 328B N-gram embedding parameters. It supports text, image, and video inputs with a 1M-token context window and up to 393K output tokens.
Background
What is an embedding? Vector representations for similarity and retrieval
Embeddings map content into numerical vectors for learned similarity comparisons, often supporting retrieval or clustering. Similarity is not factual equivalence or correctness. Use compatible models and metrics rather than mixing equal-dimensional vectors from different models. Keep source text separately; embeddings are not losslessly recoverable document archives.
What is a context window? Per-request capacity and the limits of recall
A context window limits the tokens a model can handle for a request. Instructions and conversation history can consume capacity, and models account for input, output, and reasoning tokens differently. A large window is neither permanent memory nor a guarantee of complete retrieval from long documents. Exceeding limits may cause errors or truncation.
Reading notes
The event summary comes from collected excerpts. Background passages quote existing HotAI articles to explain terms and do not establish new facts about this event. Check the original sources for product results, conditions and subsequent updates.
Original source
Sources and collection timeline
| Source | Report title | Original publication time | Collected by HotAI |
|---|---|---|---|
| marktechpost.com | Alibaba launches Qwen4-Max: 2.4T MoE, 71B activations, 1M context | 2026-11-16 17:44 | 2026-11-16 18:27 |