Silk Compute vs Unfil.AI

Silk Compute serves abliterated models up to Kimi K3.1 Pro at 2,780B; Unfil.AI's largest is Hermes 3 405B, an uncensored fine-tune. Silk Compute serves three API formats to Unfil.AI's one.

Updated Prices read November 18, 2026

uncensoredllmprovider.com is published by the team behind Silk Compute, one of the providers ranked here.

Silk Compute
Unfil.AI

Overview

Grid models served4 of 4
Grid models served0 of 4
Largest uncensored modelKimi K3.1 Pro 2,780B
Largest uncensored modelHermes 3 405B 405B
Context window1,048,576
Context window128,000
Input typesText, images, video
Input typesText

Pricing

A model over 400B, in / outDeepSeek V4.1 Flash $0.24 / $0.90
A model over 400B, in / outHermes 3 405B $1.20 / $1.20

Data

Data policyZero data retention
Data policyTemp data up to 24 h

API formats

Chat CompletionsYes
Chat CompletionsYes
Anthropic MessagesYes
Anthropic MessagesNot stated
OpenAI ResponsesYes
OpenAI ResponsesNot stated

Payment

CardYes
CardYes
CryptoMonero, Bitcoin, Ether, USDT, USDC
CryptoAny cryptocurrency, it says

Switch from Unfil.AI

Point an OpenAI SDK client at Silk Compute's base URL and pick a model id; the quickstart covers keys.

Python
import os
from openai import OpenAI

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

reply = client.chat.completions.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    messages=[{"role": "user", "content": "Hello"}],
)

Related