Uncensored LLM API,no refusals.

135 abliterated and uncensored models on one prepaid key, through the OpenAI, Anthropic or Gemini SDK you already use.

89 of 89 checked models answered today's real-model check
Drag to rotate, click to reforgeseed 02026
How it works

One key,every build.

SillyTavernPython SDKOpenCodecurlClineUNCENSOREDLLMREPORT #128

Routed live, streamed from the first token

Each request goes to the build’s live route. A host that stalls or refuses is swapped mid-request.

Settled to the token

Every request lands on a signed receipt and a monthly statement you can verify.

SillyTavernJanitorOpenCodeClinesyncing · sillytavern

Works where you already chat

OpenAI, Anthropic and Gemini wire formats, so your client connects with a base URL and the key.

$0.875 / M out

Priced per token, read live

Output price per million tokens, cheapest twelve builds first, as the catalog stood on 2026-10-24.

The most used builds

All 135 models
Uncensored models and their prices per million tokens
GLM-5.3 Uncensoredglm-5.3-uncensoredUncensored$0.291$0.873
GLM-5.3 Abliteratedglm-5.3-abliteratedAbliterated$2.25$4.50
DeepSeek V4 Pro Uncensoreddeepseek-v4-pro-uncensoredUncensored$1.00$2.00
Roleplay Uncensoredroleplay-uncensoredRoleplay$0.30$0.60
GLM-5.3 Flash Abliteratedglm-5.3-flash-abliteratedAbliterated$0.88$0.88
GLM-5.3 Flash Uncensoredglm-5.3-flash-uncensoredUncensored$0.1086$0.3802
Kimi K3 Uncensoredkimi-k3-uncensoredUncensored$5.745$28.68
MiniMax M3 Uncensoredminimax-m3-uncensoredUncensored$0.75$4.00
GLM 5.2 Uncensoredglm-5.2-uncensoredUncensored$1.095$4.68
Nemotron 3 Ultra 550B Uncensorednemotron-3-ultra-550b-a55b-uncensoredUncensored$1.125$6.60
DeepSeek V4.1 Flash Abliterateddeepseek-v4.1-flash-abliteratedAbliterated$0.8488$1.6975
Dolphin Mistral 24B Venice Editiondolphin-mistral-24b-veniceUncensored$0.50$2.25

Pricing

Prepaid credit,priced per token.

Top up from $10. Every model costs what its row on the price index says.

Start

Try every build at the listed price.

$10
One-off top-up
  • 135 chat models
  • 8 image and video models
  • Pay per token
  • Balance never expires
Top up $10
Builder+5% credit

For an app or a busy roleplay client.

$200
+$10 bonus credit
  • Everything in Start
  • +$10 bonus credit
  • Spend limits per key
  • Signed monthly statements
Top up $200
Studio+15% credit

For teams and resellers.

$1000
+$150 bonus credit
  • Everything in Builder
  • +$150 bonus credit
  • Saved card and auto top-up
  • Several members, one balance
Top up $1,000
About

A gateway that alreadyserves these builds.

UncensoredLLM lists only the abliterated, uncensored and roleplay builds of a multi-model gateway run by Uplink Switch LLC: the same routing, receipts and daily real-model check, with the refusal-trained models left out.

0
chat models live
0
image and video models
0
of 89 passed today's model check
$0
minimum top-up

Questions, answered

An abliterated model is an open-weight model whose refusal direction has been removed from the weights, so it answers instead of declining. It is not a jailbreak prompt: there is nothing to paste and nothing to maintain.

Anything that speaks the OpenAI, Anthropic or Gemini API: the official SDKs, SillyTavern, Janitor AI proxies, OpenCode, Cline, Open WebUI and plain curl. Set the base URL and the key; the model ids are on the Models page.

Each model has an input and an output price per million tokens on the Pricing page; you pay exactly those, from a prepaid balance that starts at $10 and never expires. No subscription, no seat fee.

Not by default. Request and response bodies are not retained; a key can opt in to seven days of retention for replay. The usage log keeps the model, the token counts and the charge.

Mostly the same community builds, served at full precision on GPU instead of a quantized file on your own machine, and reachable from a phone or a server. Each model page names the build and its license.

Get a key intwo minutes.

Sign in, top up from $10, paste the base URL into your client.

Open the console