Core Features
Unfiltered by default*
We don't rewrite your thoughts into 'safe style'. Illegal and harmful scenarios are blocked.
*With guardrails against harm
Streaming answers
Responses appear in real-time — we type faster than you read.
Privacy-first
Keys and routing stay server-side. History can be made ephemeral.
Session threads
Maintain multiple conversations in parallel. Clear with one click.
Prompt controls
Switch between ready-made tone profiles (Balanced / Direct / Safe) without diving into code.
Exports
Copy to Markdown, TXT, share read-only links.
Interactive Demos
Temperature Control
Temperature: 0.7
Creative, varied responses
Prompt Presets
Thoughtful, nuanced responses
Privacy Mode
Ephemeral Mode
Conversations saved locally
Performance & Reliability
~200ms
Median Latency
99.9%
Uptime
~50
Tokens/sec
Live metrics coming soon
HellGPT vs Typical AI
| Feature | HellGPT | Typical AI |
|---|---|---|
| Content filtering | Minimal* | Heavy |
| Streaming responses | ✓ | Sometimes |
| Privacy controls | Server-side keys | Varies |
| History management | Easy clear | Complex |
| Export options | ✓ | Limited |
*With guardrails against illegal and harmful content
Power-user Shortcuts
Send message⌘/Ctrl + Enter
Edit last message↑
Clear inputEsc
Clear conversation⌘/Ctrl + K
Toggle shortcuts⌘/Ctrl + /
Safety & Responsibility
No Illegal Use
We block requests for illegal activities and harmful instructions.
Harm Prevention
Content that could cause real-world harm is filtered out.
Rate Limiting
Abuse prevention through intelligent rate limiting.