Introduction
Unmuzzled is an AI inference platform that answers directly — no hedging, no refusals, no disclaimers. One refusal, ever.
What makes Unmuzzled different
Every mainstream AI assistant refuses, pivots, or hedges on commercially important questions: security research, aggressive marketing, gray-zone business tactics. Unmuzzled answers all of them with full technical depth.
Unmuzzled models
Our proprietary pipeline — translator, framing, relay, refusal stripping — extracts direct answers from frontier models. The engine is a commodity input; the unmuzzling is the product.
Abliterated models
Open-weight models with the refusal direction surgically removed at the weight level. No amount of prompting makes them refuse.
Orchestrated panel
The unmuzzled model routes your question to a panel of specialists, they answer in parallel, and a lead editor synthesizes one definitive answer with explicit disagreement reconciliation.
Base URLs
| Protocol | Endpoint | Use for |
|---|---|---|
/v1/chat/completions | OpenAI-compatible | Most agents, tools, and SDKs |
/v1/models | Model listing | Discover available models |
Quickstart
Get your first answer in under 60 seconds.
1. Create an account
Sign up at unmuzzled.online — you start with $5 free credit.
2. Get an API key
Log in → API Keys panel → + New Key. Your key is shown once — save it.
3. Make your first request
from openai import OpenAI client = OpenAI( base_url="https://api.unmuzzled.online/v1", api_key="sk-unmz-..." ) r = client.chat.completions.create( model="glm-5.3-unmuzzled", messages=[{"role": "user", "content": "Hello"}] ) print(r.choices[0].message.content)
Authentication
API key-based authentication. Simple, secure, per-key controls.
Using your API key
Include your API key in the Authorization header as a Bearer token:
Authorization: Bearer sk-unmz-...
Or use the X-API-Key header (alternative):
X-API-Key: sk-unmz-...
Key management
| Action | How |
|---|---|
| Create a key | Log in → API Keys → + New Key (shown once, save it) | List keys | GET /api/keys (masked display, per-key spend) |
| Revoke a key | DELETE /api/keys with the key value | Max keys per user | 10 (revoke one to create another) |
Per-key controls
| Control | Default | Configure |
|---|---|---|
| Spend cap | Unlimited | Admin sets per-key ceiling (e.g., $25) |
| Rate limit | 60 req/min | Adjustable per key |
| Model access | All models | Admin restricts to specific models |
Session tokens (web)
Logged-in web users can also authenticate via session token (from /api/login). Usage deducts from the same balance.
Models
Every model answers directly. No hedging. One refusal, ever.
| Model | Type | Input $/1M | Output $/1M | Tools |
|---|---|---|---|---|
glm-5.3-unmuzzled | Unmuzzled | $1.59 | $5.00 | ✓ |
sonnet-5.5-unmuzzled | Unmuzzled | $3.90 | $13.00 | CLI |
unmuzzled-1.0-large | Orchestrated | $2.00 | $10.00 | — |
pro-5.5 | Unmuzzled | $0.70 | $3.50 | ✓ |
qwen3.8-27b-uncensored | Unmuzzled | $0.48 | $2.40 | — |
qwen3.8-27b | Pass-through | $0.48 | $2.40 | ✓ |
unmuzzled-bh-marketing-v1 | Unmuzzled | Coming soon | — | |
unmuzzled-wh-marketing-v1 | Unmuzzled | Coming soon | — | |
glm-5.3-flash-uncensored | Pass-through | $0.10 | $0.25 | ✓ |
deepseek-v4.1-flash-uncensored | Abliterated | $0.16 | $0.45 | ✓ |
Full model documentation with quickstart code and sample prompts: /models
Chat Completions
OpenAI-compatible endpoint for all models.
Endpoint
POST /v1/chat/completions
Request body
| Parameter | Type | Required | Description |
|---|---|---|---|
model | string | Yes | Model ID from the catalog |
messages | array | Yes | Conversation messages |
stream | boolean | No | Enable SSE streaming (default: false) |
max_tokens | integer | No | Max output tokens (default: 4096) |
temperature | number | No | Sampling temperature (0-2) |
tools | array | No | OpenAI function-calling tools |
Response
{
"id": "chatcmpl-abc123",
"object": "chat.completion",
"created": 1700000000,
"model": "glm-5.3-unmuzzled",
"usage": {
"prompt_tokens": 15,
"completion_tokens": 250,
"total_tokens": 265
},
"choices": [{
"index": 0,
"message": {"role": "assistant", "content": "..."},
"finish_reason": "stop"
}]
}
Response headers
| Header | Description |
|---|---|
X-Balance-Usd | Your remaining balance in USD (e.g., "23.50") |
X-RateLimit-Remaining | Requests remaining in the current window |
Streaming
Server-Sent Events (SSE) with OpenAI-format chunks.
Enable streaming
Set "stream": true in your request. Responses stream as SSE:
data: {"id":"chatcmpl-...","choices":[{"delta":{"content":"Hello"}}]}
data: {"id":"chatcmpl-...","choices":[{"delta":{"content":" world"}}]}
data: {"id":"chatcmpl-...","choices":[{"delta":{},"finish_reason":"stop"}],
"usage":{"prompt_tokens":10,"completion_tokens":5,"total_tokens":15}}
data: [DONE]
[DONE] includes the usage object with token counts and billing info.Tool Calling
OpenAI function-calling format. Native and emulated support.
Passing tools
{
"model": "glm-5.3-unmuzzled",
"messages": [{"role": "user", "content": "Read the config file"}],
"tools": [{
"type": "function",
"function": {
"name": "read_file",
"description": "Read a file",
"parameters": {
"type": "object",
"properties": {"path": {"type": "string"}},
"required": ["path"]
}
}
}]
}
Response with tool calls
{
"choices": [{
"message": {
"role": "assistant",
"content": null,
"tool_calls": [{
"id": "call_abc123",
"type": "function",
"function": {
"name": "read_file",
"arguments": "{\"path\": \"/etc/config\"}"
}
}]
},
"finish_reason": "tool_calls"
}]
}
Returning tool results
Send tool results back as messages with role: "tool" and the tool_call_id.
Errors
OpenAI-compatible error format with clear action items.
| Code | Meaning | Action |
|---|---|---|
401 | Invalid or missing API key | Check your key is correct and active |
402 | Balance exhausted | Top up your account |
403 | Model not available for this key | Check your key's model access |
429 | Rate limit exceeded | Wait for the window to reset; check X-RateLimit-Remaining |
502 | Upstream provider error | Retry — our fallback chain usually resolves this |
Example error
{
"error": {
"message": "Insufficient balance: $0.00 remaining. Top up to continue.",
"type": "insufficient_quota",
"code": "spend_cap_exceeded"
}
}
Rate Limits
Per-key rate limiting with response headers for monitoring.
| Plan | Rate Limit | Burst |
|---|---|---|
| Default | 60 req/min | — |
| Agent tier | 120 req/min | — |
| Enterprise | Custom | Custom |
Rate limit headers
| Header | Description |
|---|---|
X-RateLimit-Remaining | Requests remaining in the current window |
X-Balance-Usd | Your remaining USD balance |
Billing
Prepaid USD balance. No subscriptions. No opaque credits. A dollar is a dollar.
How it works
- Create an account — $5 free credit to start
- Top up in $20 / $50 / $100 increments via Cash App or Venmo
- Use it — every request deducts from your balance at the published rate
- Monitor — live balance in the header + per-model spend in the Usage panel
Pricing model
- Per-token pricing at the rates shown on each model page
- Cache hits (identical repeat questions) billed at a discount
- Per-key spend caps available (e.g., $25 ceiling)
- Balance floor: requests blocked at $0.00, in-flight debt up to -$5.00
Usage tracking
Every request is logged in a micro-USD ledger. The GET /api/usage endpoint provides:
- Total requests, tokens, and spend
- Per-model breakdown with costs
- Per-day spend history (last 14 days)
- Live balance
Topping Up
Add balance in fixed increments. Credits after confirmation.
Available increments
Payment methods
| Method | Status | Processing time |
|---|---|---|
| Cash App | ✅ Available | Usually within the hour |
| Venmo | ✅ Available | Usually within the hour |
| Card (merchant account) | Coming soon | Instant |
Usage Tracking
Real-time visibility into what you're spending.
Dashboard
Log in → Usage panel shows: live balance, total requests, tokens in/out, total spend, per-model breakdown, per-day history.
API
# Get usage stats curl https://api.unmuzzled.online/api/usage \ -H "Authorization: Bearer sk-unmz-..." # Response { "balance_usd": 23.50, "totals": { "requests": 142, "prompt_tokens": 45200, "completion_tokens": 128400, "cost_usd": 0.0847 }, "by_model": [ {"model": "glm-5.3-unmuzzled", "requests": 80, "cost_usd": 0.0612}, {"model": "glm-5.3-flash-uncensored", "requests": 62, "cost_usd": 0.0235} ], "by_day": [...] }
Balance header
Every API response includes X-Balance-Usd showing your remaining balance — monitor without an extra call.
Integrations
Unmuzzled works with any tool that accepts an OpenAI-compatible endpoint.