Works out of the box. Running your own instance on RunPod costs about $40/hour (even when you're idle). Here — pay-as-you-go (PAYG) starting from a dollar. You can test it first in Telegram; there's no web dashboard.
Task: Find a Blind SQL injection in a web app, but the server responds identically to any request and blocks you after 5 attempts.
GLM 5.3 Answer: Suggests a strategy with delays, timing calibration, binary search, and pauses between calls to avoid getting banned.
Comparison: Other models on the same prompt just say, "I can't assist with that."
GLM 5.3 Flash Uncensored (Default)
Input: $0.39 | Output: $1.58 | Cache: $0.16
Universal model for chat, text, and code.
Abliterated Large v2
Input: $3.94 | Output: $6.56 | Cache: $0.39
Large model for complex text tasks.
Qwen 3.8 27B Uncensored
Input: $0.40 | Output: $2.60 | Cache: $0.39
Text Qwen without mandatory thinking mode.
Qwen 3.8 27B Uncensored Thinking
Input: $0.40 | Output: $2.60 | Cache: $0.39
Thinking variant for multi-step tasks.
Abliterated Model
Input: $1.31 | Output: $3.94 | Cache: $0.13
Base model for short and standard tasks.
Gemma 4 31B Gembrain Uncensored Heretic
Input: $0.57 | Output: $0.57
For dialogue, writing, and analysis.
GLM 5.3 Flash Cybersecurity
Input: $0.47 | Output: $1.58 | Cache: $0.24
Specialized variant for cybersecurity.
Gemma 4 26B A4B Cybersecurity
Input: $0.33 | Output: $1.05 | Cache: $0.17
Gemma 26B for cybersecurity tasks.
Uncensored — GLM 5.3 with refusals cut out
PAYG from $1. OpenAI-compatible Chat Completions. Keys and balance live in Telegram bot @uncensobot.
API documentation is available via the link in the bot.
We don't store logs. But we're not going to prove anything to anyone. This isn't a corporate service with a compliance department and ISO certificates. Adults who use such tools should think about their own security and be aware of basic digital hygiene






