Use case · Trust & safety
Moderate what other models won't read.
Classify, label and explain harmful content with a model that can look at it.
Built forModeration & policy teams
Why refusals get in the way
Trust and safety teams read what people actually post: abuse, extremism, scams and harassment. A model that won't engage with the content can't label it, explain which rule it breaks or help write the policy that covers it.
What teams use it for
- Triage and label reported content against your policy
- Explain decisions and draft appeal responses
- Test policy definitions on real edge cases
- Generate lawful synthetic examples to train and evaluate classifiers
What Lyber gives you
Answers, not refusals
Uncensored and abliterated weights, with no refusal layer of Lyber's own on top.
Your policy, in your system message
The scope and rules for each request travel with it, in the system message you write.
One OpenAI-compatible API
Keep your SDK, harness and agents. Change the base URL and the model name.
Keys, limits and records
A key per app or person with spending limits, shared organizations, and a usage ledger designed to exclude prompt text.
The request
POST /v1/chat/completions
{
"model": "lyber/auto",
"messages": [
{ "role": "system", "content": "<your content policy and labels>" },
{ "role": "user", "content": "<your request>" }
]
}Early access. Inference opens as each model is qualified, and launch prices are provisional.