Skip to content
Models · Aug 26, 2026

Z.ai confirms Ox Alpha as its new open-weight reasoning model with weights due August 28

The GLM-series developer says Ox Alpha tops benchmarks and will ship with weights on Wednesday, positioning it against frontier U.S. labs.

Trust74
HypeLow hype

1 source · cross-referenced

ShareXLinkedInEmail
TL;DR
  • Z.ai confirmed it built Ox Alpha, a new open-weight reasoning model that currently leads public benchmarks.
  • The company says Ox Alpha is designed for coding, long-horizon agentic work, and production workloads with text and visual inputs.
  • Weights for Ox Alpha are scheduled for release on August 28, enabling developers to build on the model.
  • Z.ai describes Ox Alpha as the latest iteration in its GLM series, following the recent GLM-5.3 release.

Z.ai confirmed it is the developer of Ox Alpha, a previously anonymous open-weight AI model that had rapidly risen to the top of public benchmarks and leaderboards. The company stated Ox Alpha is the newest iteration of its GLM series.

Z.ai characterizes Ox Alpha as a reasoning model intended for coding, sustained agentic work, and production workloads, and says it supports workflows that combine text with visual context. The company did not provide specific benchmark figures in its announcement.

The lab said it will release the model weights on August 28, enabling developers to fine-tune and deploy the model. Z.ai did not specify the model size or training compute in its announcement.

Z.ai also noted that Ox Alpha follows the recent release of GLM-5.3, which the company said rivals Anthropic’s Fable 5 on certain benchmarks.

Sources
  1. 01TechCrunch — AISurprise: Z.ai is the AI lab behind the mysterious Ox Alpha model
Also on Models

Stories may contain errors. Dispatch is assembled with AI assistance and curated by human editors; despite the trust-score filter, mistakes happen. We correct publicly — every article links to its revision history. Nothing here is financial, legal, or medical advice. Verify before relying on any claim.

© 2026 Dispatch. No ads. No sponsorships. No paid placement. Reader-supported via Ko-fi.

Built by a person who cares about honest AI news.