Updated

Abliterated model with the Anthropic SDK

The Anthropic SDKs reach an abliterated model through POST /v1/messages once you pass the provider's base URL. The SDK adds /v1/messages itself, so pass the bare base URL.

  1. 1

    Install the SDK

    Both SDKs read the base URL from the client options.

    Pythonpip install anthropic
    TypeScriptnpm install @anthropic-ai/sdk
  2. 2

    Set your key

    Create a key on Silk Compute's API keys page and add credits from $5. Keep it in an environment variable.

    Shellexport SILK_API_KEY="sk-..."
  3. 3

    Send a message

    Without thinking the model thinks as little as it can. A budget under 8,192 runs as low effort, and 32,768 or more as max.

https://api.silkcompute.com
import os
from anthropic import Anthropic

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

message = client.messages.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    max_tokens=2048,
    messages=[{"role": "user", "content": "What is a mutex?"}],
)
print(message.content[0].text)

Example output

A lock only one thread can hold at a time.
import Anthropic from '@anthropic-ai/sdk'

const client = new Anthropic({
  baseURL: 'https://api.silkcompute.com',
  apiKey: process.env.SILK_API_KEY,
})

const message = await client.messages.create({
  model: 'deepseek-v4.1-flash-1017-uncensored',
  max_tokens: 2048,
  messages: [{ role: 'user', content: 'What is a mutex?' }],
})
console.log(message.content)

Example output

[ { type: 'text', text: 'A lock only one thread can hold at a time.' } ]
message = client.messages.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    max_tokens=16000,
    thinking={"type": "enabled", "budget_tokens": 8192},
    messages=[{"role": "user", "content": "Find the race condition in this function."}],
)

for block in message.content:
    if block.type == "thinking":
        print("thinking:", block.thinking)
    elif block.type == "text":
        print(block.text)
const message = await client.messages.create({
  model: 'deepseek-v4.1-flash-1017-uncensored',
  max_tokens: 16000,
  thinking: { type: 'enabled', budget_tokens: 8192 },
  messages: [{ role: 'user', content: 'Find the race condition in this function.' }],
})

for (const block of message.content) {
  if (block.type === 'thinking') console.log('thinking:', block.thinking)
  if (block.type === 'text') console.log(block.text)
}

What differs from Anthropic's API

The full parameter list is in the Messages reference.

FeatureOn /v1/messages
Video inputReturns 400
PDFs and token countingNot supported
Server tools such as web searchNot available
cache_control markersNot needed, caching is automatic
Error shape{"type": "error", "error": {...}}