Qwen 3.8 27B — Uncensored

Qwen 3.8 27B on our own hardware, used for the 10 free replies. API and credit-funded Chat can use a compatible backup when the primary is unavailable.

Input / 1M tokens$0.20

Cached input$0.20

Output / 1M tokens$0.60

Context length262,144

StatusAvailable

Model informationOpen chat API Playground Docs

Model and API details

Model version: Qwen 3.8 27B — UncensoredThe exact weight revision is not published.
Hosting: Own hardware with a backup route
Refusal reduction: A Qwen version tuned to reduce refusals.
API interface: Chat Completions · SSEUse the Chat Completions examples below. Tool calls, image input and other OpenAI endpoints need separate compatibility checks.
Context length: 262,144 tokensCurrent serving configuration.
Maximum output tokens: 32,768 tokensInput and output share the context window. The free reply trial has its own smaller limits.

The primary route is preferred. A compatible backup may serve requests when it is unavailable. Prompt caches are separate across routes; cached-input prices apply only to reported cache hits.

Compare answers on your own tasks before choosing. These are configured or provider-published specifications; independent quality and maximum-context results have not yet been published.

Information reviewed: 28 October 2026.

A familiar interface

Use an OpenAI-compatible client with your API key. Select a model ID from the live model directory.

JavaScript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://uncensoredai.click/v1",
  apiKey: process.env.UNCENSOREDAI_API_KEY,
});

const response = await client.chat.completions.create({
  model: "Qwen_3.8_27B_Uncensored",
  messages: [{ role: "user", content: "Hello!" }],
});