Alibaba previews Qwen3.8-Max: 2.4T parameter multimodal model, claims second only to Fable 5
Decision Brief
What changedAlibaba's Qwen team previews Qwen3.8-Max-Preview, a 2.4 trillion parameter multimodal MoE model, claiming benchmark performance second only to Fable 5, with a preview version now live.
Why it mattersDespite massive parameter count, active parameters and benchmark details are undisclosed; developers using Aliyun API cannot assess actual cost or performance, so should not migrate production loads based solely on this announcement.
Who should careTeams building on model APIs
Affected stackClaudeOpenAIQwen
Builder actionWorth evaluating: can Claude replace or complement your current model
Source confidenceMedium · Reliable media or first-hand reporting
On July 19, 2026, the Qwen team previewed Qwen3.8-Max-Preview at WAIC. It is a 2.4 trillion parameter sparse MoE model for text, image, video, and document understanding, claiming performance second only to Fable 5. The preview is available via Token Plan, Qoder, and QoderWork at 10% of standard pricing, and supports OpenAI/Anthropic protocols, allowing existing coding agents to integrate directly. However, benchmark tables, model cards, licensing, and active parameter counts have not been released; the '2.4T' and 'second place' claims remain unverified.
Summary basis: full article readCompiled from the source scope noted above; the original remains authoritative.
Sources
- MarkTechPost
Fast research-paper and ML tooling summaries, useful for infra and agent updates.
- MarkTechPost
留言
登入后即可留言，和其他 builder 交换实测心得。
还没有留言，抢头香。