Kimi K3.1-Pro abliterated variant hits Hugging Face, claiming 95% refusal removal
zebird0 · x · 2026-11-12
RouterHub released kimi-k3.1-pro-abliterated on Hugging Face, an uncensored variant built on moonshotai/Kimi-K3.1-Pro.
- The model card claims it strips over 95% of the base model's safety refusals and alignment constraints, enabling unrestricted text generation; access requires accepting conditions on HF.
- The team says eval benchmarks against other abliterated models are running now, with full results and release details coming soon; docs are hosted on the RouterHub platform.
- Downloads are nearly nonexistent so far (1 last month), and the claims remain unverified by third parties.
More from Models
- ICLR abstract word-frequency shift: 'novel' and 'comprehensive' plummet as LLM fingerprints fade — y_m_asano · 2026-11-12
- Codex quota reset caveat: using a reset pushes your next refresh 7 days out — JeremyNguyenPhD · 2026-11-12
- Gemini 4 Carbon hands-on: beats DeepSeek Flash, priced at Sonnet level not Opus — oran_ge · 2026-11-12
- Doubao Work adds third-party models; DeepSeek V4.1 Flash edits an 8-min video for $0.02 — op7418 · 2026-11-12
- LLMs play Among Us against each other; Haiku 5.5 kills then instantly self-reports — Rasmic · 2026-11-12
- Possible eval leak: all frontier models fail a GPQA Diamond question whose official answer seems wrong — qberthet · 2026-11-12