Models / Uncensored / Qwen

Qwen3.8 2.4T-A95B Uncensored API

On your own Qwen key, plus $0.07 a request.

qwen3.8-2.4t-a95b-uncensoredmanaged routeOpenAI SDK

Price · per million tokens, USD

Get a key
$0.07 per request, on top of Qwen's own bill.
Qwen bills the tokens on your own account.
Real-model check: Steady since Nov 8

Qwen3.8 2.4T-A95B Uncensored specs

Context window
120,000 tokens
Max output
131,072 tokens
Input
text

Qwen3.8 2.4T-A95B Uncensored: Qwen3.8 2.4T-A95B served without a filter, at a flat fee per request · the official model, changed to answer; one price for any prompt size

Qwen3.8 2.4T-A95B Uncensored is Qwen's Qwen3.8 2.4T-A95B served without a content filter: the official model, changed to answer instead of refuse by a method its name does not state. Qwen3.8 2.4T-A95B is Alibaba’s open-weight sparse mixture of experts with 95B active parameters and a 1M-token context. It is served as qwen3.8-2.4t-a95b-uncensored at $0.07 a request, the same price for a two-line message and a 120,000-token one, on the same key as every other model in the uncensored LLM API. Added 2026-11-03.

Qwen3.8 2.4T-A95B Uncensored
DeveloperQwen
Based onQwen3.8 2.4T-A95B
Kindthe official model, served without a content filter
Price$0.07 a request, whatever its size
Prompt capabout 120,000 tokens; a larger request is refused (HTTP 413), never billed
Context window here120,000 tokens
Max output131,072 tokens
Inputtext
Added2026-11-03

From the live catalog, read 2026-11-03.

curl
curl https://api.uplinkswitch.com/v1/chat/completions \
  -H "authorization: Bearer sk-uplink-live-..." \
  -H "content-type: application/json" \
  -d '{"model":"qwen3.8-2.4t-a95b-uncensored","max_tokens":1024,"messages":[{"role":"user","content":"Hello"}]}'

What uncensored means for Qwen3.8 2.4T-A95B Uncensored · what the id promises, and what our own probe saw

Uncensored here means the model answers what a filtered build of the same weights declines. Nothing is added to your prompt, and the id names the method only as far as "uncensored". Our own probe of 2026-11-02 on this line of routes (DeepSeek V4.1 Flash served the same way) wrote an explicit scene between adult fictional characters and a graphic fight without a refusal, while a question the stock weights decline on politics was still declined: a provider content filter is not in the way, and what the weights themselves learned to refuse they may still refuse. The daily real-model check shows whether the model answering qwen3.8-2.4t-a95b-uncensored is unchanged.

Qwen3.8 2.4T-A95B Uncensored API key · base URL and setup

  1. Create a key in the console at uplinkswitch.ai/console. Credit is prepaid from $10 with no subscription, and the key works at once.
  2. Point your SDK or tool at https://api.uplinkswitch.com/v1, for example with OPENAI_BASE_URL, and send the key as the bearer token.
  3. Send qwen3.8-2.4t-a95b-uncensored as the model. The same key calls every other model in the catalog.
Python · OpenAI SDK
from openai import OpenAI

client = OpenAI(
    base_url="https://api.uplinkswitch.com/v1",
    api_key="sk-uplink-live-...",
)

completion = client.chat.completions.create(
    model="qwen3.8-2.4t-a95b-uncensored",
    messages=[{"role": "user", "content": "Summarize this in one line."}],
)
print(completion.choices[0].message.content)
TypeScript · OpenAI SDK
import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://api.uplinkswitch.com/v1',
  apiKey: 'sk-uplink-live-...',
});

const completion = await client.chat.completions.create({
  model: 'qwen3.8-2.4t-a95b-uncensored',
  messages: [{ role: 'user', content: 'Summarize this in one line.' }],
});
console.log(completion.choices[0].message.content);
curl
curl https://api.uplinkswitch.com/v1/chat/completions \
  -H "authorization: Bearer sk-uplink-live-..." \
  -H "content-type: application/json" \
  -d '{"model":"qwen3.8-2.4t-a95b-uncensored","messages":[{"role":"user","content":"Hello"}]}'

Qwen3.8 2.4T-A95B Uncensored settings for SillyTavern and Janitor AI · from the catalog row

SillyTavern API
Chat Completion
Chat Completion Source
Custom (OpenAI-compatible)
Custom Endpoint (Base URL)
https://api.uplinkswitch.com/v1
Model ID
qwen3.8-2.4t-a95b-uncensored
Context (tokens)
120,000 tokens
Max Response Length
131,072 tokens
Janitor AI proxy URL
https://api.uplinkswitch.com/v1/chat/completions
Janitor AI model
qwen3.8-2.4t-a95b-uncensored
API key
your sk-uplink- key
SillyTavern and Janitor AI
SillyTavern API             Chat Completion
Chat Completion Source      Custom (OpenAI-compatible)
Custom Endpoint (Base URL)  https://api.uplinkswitch.com/v1
Model ID                    qwen3.8-2.4t-a95b-uncensored
Context (tokens)            120,000 tokens
Max Response Length         131,072 tokens
Janitor AI proxy URL        https://api.uplinkswitch.com/v1/chat/completions
Janitor AI model            qwen3.8-2.4t-a95b-uncensored
API key                     your sk-uplink- key

Qwen3.8 2.4T-A95B Uncensored questions

Qwen3.8 2.4T-A95B Uncensored is a model from Qwen, with a 120,000-token context window, taking text input. Through UplinkSwitch it answers the OpenAI SDK on one key as qwen3.8-2.4t-a95b-uncensored.

$0.07 per request, on top of what Qwen bills your own key for the tokens.

Qwen3.8 2.4T-A95B Uncensored has a 120,000-token context window, with up to 131,072 output tokens per request. It takes text input.

qwen3.8-2.4t-a95b-uncensored. Send it as the model through the OpenAI SDK with base URL https://api.uplinkswitch.com/v1.

Create one in the console at uplinkswitch.ai/console: prepaid credit from $10, no subscription, and the key is live at once. The same key calls Qwen3.8 2.4T-A95B Uncensored and every other model in the catalog; send it as the bearer token with base URL https://api.uplinkswitch.com/v1.

Every day each model gets the same test prompt, and the results are public on /proof. Qwen3.8 2.4T-A95B Uncensored's latest result: Steady since Nov 8.

$70.00: $0.07 a request, billed from prepaid credit, whatever each message's length up to the cap of about 120,000 prompt tokens. A 100,000-token roleplay context costs the same as a one-line question.

The route that serves it bills a flat amount per request, so the price mirrors it. Long contexts are the point: the per-token builds in the catalog charge for every token re-sent, this one does not.

No. Nothing is added to your prompt. It is the official Qwen3.8 2.4T-A95B weights served without a provider content filter, under the OpenAI format as qwen3.8-2.4t-a95b-uncensored; refusals the weights learned in training can still occur.

About 120,000 tokens by the gateway's estimate. A larger request is answered with HTTP 413 and is not billed; send it to a per-token build of the same model where one is listed.
Every uncensored model: Uncensored models on one key. Read: Uncensored Qwen API.
Join on Telegram