# Alibaba's Qwen3.8-Max Posts Agentic Benchmark Scores Within Points of Closed Frontier Models

The 2.4-trillion-parameter sparse model activates 95 billion parameters per token and lists at $2 and $6 per million input and output tokens, well below comparable closed-model rates.

- Published: 2026-08-16T06:12:17.236Z
- Canonical: https://polylog.news/ai/2026-08-16/alibaba-s-qwen3-8-max-posts-agentic-benchmark-scores-within
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [MarkTechPost](https://www.marktechpost.com/2026/08/03/alibaba-qwen-releases-qwen3-8-max/), [MindStudio](https://www.mindstudio.ai/blog/qwen3-8-2-4t-a95b-model-overview), [Digital Applied](https://www.digitalapplied.com/blog/qwen3-8-max-full-release-benchmarks-open-weights), [Developers Digest](https://www.developersdigest.tech/blog/qwen-3-8-max-release-2026)

Alibaba's Qwen team released Qwen3.8-Max on August 3, a mixture-of-experts (MoE) model with 2.4 trillion total parameters and 95 billion active per forward pass. It natively handles 262,144 tokens of context, extensible to roughly one milli…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-08-16/alibaba-s-qwen3-8-max-posts-agentic-benchmark-scores-within (subscription information: https://polylog.news/pricing).