Decoding OpenCode's Mystery Model: 600 Calls Confirm It as GLM-5
zainhas · x · 2026-08-23
The author investigated OpenCode's mysterious model Ox Alpha using an autonomous Qwen red-team agent, executing over 600 API calls and consuming 13.5M prompt tokens. By analyzing technical fingerprints like tokenizer behavior, upstream error leaks, and context window limits, the author confirmed the model is actually a Z.ai/Zhipu AI GLM-5 series model, likely optimized in an FP8 or Turbo configuration.
Related event: Forensics Point to Zhipu's Unreleased GLM Behind Mystery Model Ox Alpha(31 posts)→
More from Models
- Unredacted Anthropic filing exposes internal data-quality debate months before Opus 3 cutoff — zacharynado · 2026-10-15
- Google says teams across the company are using Gemini 4 Argon on its hardest engineering problems — CurieuxExplorer · 2026-10-15
- Perplexity's new embedding models: index with a 9B model, search with 0.6B — antoine_chaffin · 2026-10-15
- 797GB of n-gram counts in CPU RAM let 1B models match 57% more compute — PangWeiKoh · 2026-10-15
- Leak: Moonshot's Kimi flagship 'x4' done with pre-training, K3.1 in internal testing — lyraxana · 2026-10-15
- Unsloth ships GGUF quants for Nemotron 3.5: Bolt-74B runs in 24GB RAM — danielhanchen · 2026-10-15