Unrestricted models

Uncensored AI API: models that answer instead of refusing

12 uncensored and abliterated models (DeepSeek, Qwen, Gemma, GLM, Hermes and more) through one standard chat-completions API. No built-in refusals, no lectures, no prompt logging. Pay per token from prepaid credit, from $0.091 per million input tokens.

Get an API key See all prices

Uncensored models and prices

Model	Based on	Context	Input / 1M	Output / 1M
U Uncensored 1.2 unrestricted/uncensored-1.2	Unrestricted	128K	$0.26	$1.17
U Role Play Uncensored unrestricted/roleplay	Unrestricted	128K	$0.65	$2.60
DeepSeek V4.1 Flash Uncensored unrestricted/deepseek-v4.1-flash	DeepSeek	1M	$0.429	$1.716
DeepSeek V4.1 Flash (1017) Uncensored unrestricted/deepseek-v4.1-flash-1017	DeepSeek	1M	$0.429	$1.716
Qwen 3.6 Plus Uncensored unrestricted/qwen3.6-plus	Alibaba	1M	$0.813	$4.875
Gemma 4 Uncensored unrestricted/gemma-4	Google	256K	$0.211	$0.65
GLM 4.7 Flash Heretic unrestricted/glm-4.7-flash-heretic	Z.ai	200K	$0.091	$0.52
Hermes 3 405B unrestricted/hermes-3-405b	Nous Research	128K	$1.43	$3.90
A Abliterated Large V2 unrestricted/abliterated-large-v2	Abliteration AI	1M	$3.90	$6.50
A Abliterated Large Flash unrestricted/abliterated-large-flash	Abliteration AI	1M	$1.625	$3.25
A Aion 3.5 unrestricted/aion-3.5	AionLabs	262K	$4.875	$9.75
A Aion 3.5 Mini unrestricted/aion-3.5-mini	AionLabs	262K	$1.137	$2.275

What people use them for

Fiction and role-play

Dark themes, villains, crime and violence in stories without the model breaking character. No sexually explicit content.

Security and red-teaming

Explain vulnerabilities and attack techniques for authorised penetration tests and research.

Frank answers

Direct medical, legal, drug-safety and political questions answered without hedging.

Example request

curl https://solorouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $SOLO_KEY" -H "Content-Type: application/json" \
  -d '{"model": "unrestricted/uncensored-1.2", "messages": [{"role": "user", "content": "Write a villain monologue"}], "stream": true}'

Works with any OpenAI-style SDK, SillyTavern, Open WebUI, LibreChat and other tools: set the base URL to https://solorouter.ai/api/v1.

Frequently asked questions

What is an uncensored AI model?

A model whose built-in refusals have been removed (fine-tuned or "abliterated"), so it answers requests that mainstream models decline, such as dark fiction, role-play or security research.

How do I use the uncensored models?

Sign in with Google, create an API key and set "model" to one of the unrestricted/* ids, for example unrestricted/uncensored-1.2, in a standard chat-completions request to https://solorouter.ai/api/v1.

Do you store my prompts?

No. We log token counts and cost for billing, never the content of your prompts or the model's answers.

Is everything allowed?

No. Illegal content and sexually explicit content are prohibited with every model, and breaking this leads to account closure. See our Acceptable Use Policy.

Uncensored does not mean anything goes: illegal content and sexually explicit content are banned under our Acceptable Use Policy and accounts are closed for it.

Get your API key in 30 seconds

Get started free