routerhubSI / uncensored-vs-abliterated-ai
git clone https://github.com/routerhubSI/uncensored-vs-abliterated-ai.git
Compares uncensored and abliterated AI model variants on RouterHub, covering model IDs, prices, refusal behavior and cost.
Why you might care
Explains the difference between abliterated (weight-level refusal reduction) and uncensored (broader product outcome) models, with exact RouterHub IDs and published prices for DeepSeek, GLM and Qwen models. Useful if you are choosing between these variants through an OpenAI-compatible API and want to compare cost; the README notes that refusal behavior must be tested per model ID.
+0
stars, 24h · #68,448
+50
stars, 7 days · #6,100
+50
stars, 30 days · #15,987
50
★ total · +50% this week
11
forks · 0 open issues
Star history (as recorded here)
LicenseMIT
Created2026-10-29 (4d ago)
Last push4d ago
Watchers2
OwnerOrganization
GitHub topics
abliterated-aiabliterated-ai-modelsai-apiai-model-comparisondeepseek-v4-1-flash-abliterateddeepseek-v4-1-flash-uncensoreddeepseek-v4-pro-abliterateddeepseek-v4-pro-uncensoredglm-5-3-abliteratedglm-5-3-flash-abliteratedglm-5-3-flash-uncensoredglm-5-3-uncensoredllm-comparisonopenai-compatiblepay-per-tokenqwen3-8-max-uncensoredrouterhubuncensored-aiuncensored-ai-modelsuncensored-vs-abliterated
Links
- GitHub: github.com/routerhubSI/uncensored-vs-abliterated-ai
- Homepage: https://routerhub.si/?utm_source=github&utm_medium=repo&utm_campaign=uncensored-vs-abliterated-ai
- Releases · Issues · Hacker News mentions
README
Loading README…
Blurb generated by claude-haiku-5-5 4d ago. Star deltas marked ~ are extrapolated from partial history.