Leaked leaderboard blur hints at three unannounced GLM models from Z.ai
Altruistic_Heat_9531 · reddit · 2026-11-15
A Reddit user deployed a DeepResearch swarm agent to chase down a leaderboard screenshot showing three blurred model names scoring 59.4% / 57.1% / 54.0%, widely believed to be unannounced GLM models.
- A community experiment reconstructed the suspected text, re-applied Gaussian blur (σx ≈ 6.9px), and got an RMSE of just 1.254/255 against the original — guessed names: GLM-Nano (or GLM-Name) Pro / Med / Small
- AICodeKing's track record: previous GLM claims landed 0-11 days before official releases (GLM-4.6 same day, GLM-5 1 day, GLM-5.1 11 days, GLM-5.2 3 days, GLM-5.3 4 days)
- The leaderboard surfaced Nov 12, near Anthropic's Nov 18 IPO; a Z.ai release this week would make for a striking coincidence
- Hugging Face activity picked up in the last 3-4 days vs the prior 27-day lull — a supporting signal, not proof
Unconfirmed speculation; no official announcement yet.
More from Models
- OpenAI weekly: ChatGPT Wallet rollout, Odessia acquisition, voice modes retiring Dec 11 — btibor91 · 2026-11-16
- AI-built 367K-parameter LM fits on a floppy disk and runs on a Pentium — RichmanRonald · 2026-11-16
- Inference providers are not hosting Perplexity's decider model despite production demand — bo_wangbo · 2026-11-16
- Science Prediction Universe benchmark: best agents score just 0.486 at predicting wet-lab outcomes — liangchen5518 · 2026-11-16
- 9 fake software brands, 18 queries to ChatGPT and Google AI Mode: zero hallucinations — raja314 · 2026-11-16
- Dean Ball: top AI models mode-collapse into 2020-era governance discourse — deanwball · 2026-11-16