Summary
The report says Ox Alpha completed all 113 tasks and was available through OpenRouter, Venice and Nous Research. The result is a researcher-reported evaluation rather than an independent audit, and the model’s underlying identity remains disputed.
Why it matters
A full-run result is stronger evidence than a limited benchmark sample, but it also highlights the need for reproducible evaluations of anonymous frontier models.
Related coverage
- Token fingerprints link anonymous Ox Alpha to Baseten’s GLM-5.2 Vision
- Waymo opens Ojai robotaxis and Gemini cabin assistant to riders in three US cities
- Marin begins a publicly tracked 535B-parameter open training run