Lyber
Request access

Models · Preparing launch

Uncensored and abliterated AI models.

Five models from four leading families, with modified uncensored or abliterated weights only. Pick a named checkpoint or send to lyber/auto. Every answer tells you which model produced it.

USD per 1 million tokens · fresh input / output

Inference is unavailable until each model and its hardware are qualified. Launch targets are provisional; the API's published price version governs billing.

Full models & pricing →Connect a client

New at Lyber

Built with the model creators.

We work directly with dealign.ai’s model-building team and pay them for model development. Follow the exact checkpoints we’re qualifying for Lyber.

Uncensored derivative

Lyber’s MiniMax M3.1 Flash

In qualification · inference unavailable

Our uncensored derivative of MiniMaxAI’s MiniMax M3.1 Flash, published under lyber-ai. Runtime, tools, cache behavior and completed-task quality are being qualified. Hugging Face weight access is gated separately.

lyber-ai/MiniMax-M3.1-Flash-UNCENSORED-NVFP4

Pricing will follow serving qualification.

View checkpoint — Lyber’s MiniMax M3.1 Flash ↗

Uncensored checkpoint

dealign.ai DeepSeek V4.1 Flash 1017

In qualification · inference unavailable

We’re qualifying dealign.ai’s uncensored 1017 checkpoint for long-context agent work. The checkpoint and its revision stay explicit; a base model or an older uncensored release is a different offering.

dealignai/DeepSeek-V4.1-Flash-1017-UNCENSORED-FP8

Provisional token targets appear in Models & pricing.

View checkpoint — dealign.ai DeepSeek V4.1 Flash 1017 ↗