// OpenAI-compatible endpoint
Unchained
LLM Access_
Uncensored and abliterated models. Auto-discovered from HuggingFace. Serverless GPU. Pay per token. No restrictions.
Models
50+
Latency
<2s cold start
Uptime
99.9%
Cost
$0/idle
// Core features
PROTOCOL
OpenAI Compatible
Drop-in replacement. Use existing SDKs and tools. Change the base URL, keep your code.
AUTOMATION
Auto-Discovery
Scout agent monitors HuggingFace around the clock. New uncensored models deployed automatically.
BILLING
Pay Per Token
Serverless GPU. No idle costs. Starting at $0.39 per 1M output tokens for 7B models.
// Request flow
$ curl api.unchained.ai/v1/chat/completions \
-H "Authorization: Bearer sk-unch-..." \
-d '{"model": "glm-4-heretic", ...}'
[1] Validate API key [2] Check credits [3] Rate limit [4] Proxy to RunPod vLLM [5] Stream response [6] Log usage
// Integration
Two lines to switch.
from openai import OpenAI
client = OpenAI(
api_key="sk-unch-your-key-here",
base_url="https://llm.ai-vfx.com/api/v1"
)
response = client.chat.completions.create(
model="davidau-qwen3-30b-a3b-claude-4-5-opus-high-reasoning-2507-abliterated-uncensored-v2",
messages=[
{"role": "user", "content": "Hello, how are you?"}
],
temperature=0.7,
max_tokens=512,
)
print(response.choices[0].message.content)// Pricing
Pay As You Go
$5+
Prepaid credits. Any model.
Pro
$49/mo
25M tokens. All models. Priority.
Business
$149/mo
100M tokens. Priority queues.