Alibaba Qwen3.8-Max debuts on Nvidia GB300: 4,000 tok/s GPU and $2M input pricing

Published: 2026-08-12 18:36:58 |

Alibaba’s Qwen3.8-Max launched Aug. 3 on Nvidia’s GB300 NVL72 inference platform, delivering over 4,000 tokens per second per GPU. The model is a 2.4T-parameter sparse Mixture-of-Experts system, with about 95B parameters active per token, and a 1 million token context window. On OSWorld-Verified, Qwen3.8-Max scored 86.1, placing it ahead of Anthropic’s Claude Fable 5 (85.0) and ahead of OpenAI’s GPT-5.6 Sol Max in the reported results. Alibaba also highlights agentic workflows, long-horizon planning, and advanced coding use cases. Pricing is set at $2 per million input tokens and $6 per million output tokens. Alibaba plans to release open weights for Qwen3.8-Max and a smaller Qwen3.8-27B variant in the week after launch, targeting researchers and developers who want to fine-tune and deploy Qwen3.8-Max directly. Nvidia’s GB300 NVL72 is also expected to reach 4,500+ tok/s per GPU for Qwen family models under optimized conditions.

Neutral

This is primarily an AI infrastructure and model-distribution news item, not a direct crypto protocol or token-specific catalyst. For traders, the connection is indirect: faster inference, open weights, and competitive API pricing (Qwen3.8-Max) can boost AI platform activity and developer adoption, but it does not directly change demand, tokenomics, or regulation for specific cryptocurrencies in the article. In the short term, the news is more likely to affect sentiment around AI/compute suppliers than crypto market stability—similar to past “model release + hardware benchmark” announcements that typically moved tech headlines without causing sustained crypto repricing. In the long term, if open-weight distribution and cost advantages lower barriers for AI agents and tooling, it could indirectly support broader ecosystem growth (and therefore certain adoptions), but the timeline and linkage to on-chain markets remain uncertain. Overall, with no named crypto assets or on-chain changes, the most reasonable read for traders is neutral: watch AI/compute equities and developer ecosystem signals, but expect limited immediate impact on crypto price action.