Uncensored LLM for trust and safety

To catch scams, abuse and harassment, you need examples of them. These models write and label that content, so you can train and test the systems that remove it.

Why refusals get in the way

Filtered models refuse to produce the very content a classifier has to learn from, or to label it accurately. Abliterated models generate realistic examples and describe them plainly.

Which model to use

GLM-5.3 Flash for high volume, at $0.30 per 1M input tokens. DeepSeek V4.1 Flash when labels need careful reasoning.

Example

Python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.silkcompute.com",
    api_key=os.environ["SILK_API_KEY"],
)

response = client.chat.completions.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    messages=[{"role": "user", "content": """Write 20 realistic scam messages aimed at elderly users, each labeled with the scam type, as training data for our detection classifier."""}],
)

print(response.choices[0].message.content)

In a coding agent

To work in a coding agent instead, set it up for Claude Code, Codex, or OpenCode. Every model works with tool calling, so the agent can read files and run commands as usual.