A mysterious new model just took over the internet...
Ox Alpha emerged as a wildly popular anonymous model on OpenRouter, later revealed as Zhipu’s GLM 5.3 Flash, impressing developers with huge context, multimodal abilities, cheap pricing, and solid coding performance despite being slow and occasionally verbose.
MAIN POINTS FROM TRANSCRIPT
- Ox Alpha became OpenRouter’s most-used model within days, handling 42 trillion tokens in six days.
- Developers suspected Chinese origins from API traces, tokenizer fingerprints, and prior anonymous animal-model patterns.
- Zhipu revealed Ox Alpha as GLM 5.3 Flash, a 320B-parameter multimodal MoE model with MIT-licensed weights.
- Its low cost, strong benchmarks, and coding ability made it attractive, though it can be slow and loop.
TAKEAWAYS
- Anonymous model launches can rapidly dominate usage when performance and price are compelling.
- Technical fingerprints often expose a model’s likely origin before official confirmation arrives.
- Cheap inference and large context windows can outweigh privacy concerns for many developers.
- Real-world coding tests suggest the model is genuinely useful, not just benchmark hype.