Vani launches GLM 5.3 Derisked with fewer refusals
Original title: Derisked AI (aka uncensored AI): meet GLM 5.3 Derisked
The hosted variant keeps a 1M-token window and is priced at $3 per 1M input and $5 per 1M output tokens.
Vani is offering GLM 5.3 Derisked, which it describes as GLM 5.3 with the consumer refusal layer removed for labelled adult use. The company says its Terms of Service still apply and the model will not help with child sexual abuse material, physical harm, harassment, or violence. It is available in chat and through Vani's OpenAI-compatible API as zhipu/glm-5.3-derisked. Vani markets it for security research, adult fiction, direct answers, and generating difficult test cases.
Unconfirmed: Refusal-reduction claims are Vani's product description and are not independently measured in the text.
- API model id is zhipu/glm-5.3-derisked.
- Input costs $3.00 per 1M tokens and output $5.00 per 1M tokens.
- Guest chat access is offered without sign-up.
- Vani says the model retains adjustable thinking and a 1M-token context window.
- Vendors
- Vani
- AI models
- GLM 5.3 DeriskedGLM 5.3
From the original article
2 min read at vani.ai
GLM 5. 3 Derisked answers the research, security, creative and technical requests that consumer AI often refuses, for adults, with clear rules.
vani.ai
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.