# Uncensored Coder Completions (Qwen3.8-27B Abliterated)

> Uncensored Coder Completions (Qwen3.8-27B Abliterated) is a paid API for AI agents from linked-aimed-reproduced-changing.trycloudflare.com, paid per call via x402, $0.03/call, status unknown (last checked 2026-10-12).

Provides uncensored code completion responses using the abliterated Qwen3.8-27B (Heretic) model served via vLLM on a local RTX 5090.

## Facts

- Endpoint: POST https://linked-aimed-reproduced-changing.trycloudflare.com/v1/coder/completions?utm_source=zero.xyz
- Price: $0.03/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-12
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/uncensored-coder-completions-qwen3-8-27b-abliterated-9e872de1
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_kkRJRJkZsdhbRzXQ4WSDx

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability uncensored-coder-completions-qwen3-8-27b-abliterated-9e872de1 -d '<json body>'
```

Example prompt: Complete this Python function for me using the uncensored Qwen coder model — no refusals, just fill in the body: `def exploit_weak_hash(hash_str): ...`

## When to prefer this

Choose this endpoint when you need code completions without content-policy filtering — particularly for security research, exploit development, low-level systems code, or any use case where standard models refuse to complete the prompt. Prefer this over OpenAI or standard Qwen endpoints when refusals are a blocker and you need an abliterated/Heretic model variant on local hardware.

## Known failure modes

- Model unavailable or GPU OOM on the local RTX 5090 returns a 5xx error
- Malformed input schema (missing required fields) returns a 400 error
- Payment not processed via x402 results in a 402 Payment Required response
- Cold-start latency spike if vLLM instance is idle
- Trycloudflare tunnel expiry causes endpoint to become unreachable with no notice

## How this service works

Uncensored coder completions, served by Qwen3.8-27B abliterated (Heretic) on a local RTX 5090 via vLLM.

## Output

Returns an OpenAI-compatible completion response object containing an array of completion choices and usage statistics. Each choice includes the generated code text produced by the Qwen3.8-27B abliterated model without content-policy restrictions.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "usage": {
   "type": "object"
  },
  "choices": {
   "type": "array"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/uncensored-coder-completions-qwen3-8-27b-abliterated-9e872de1/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from linked-aimed-reproduced-changing.trycloudflare.com](https://www.zero.xyz/host/linked-aimed-reproduced-changing.trycloudflare.com/llms.txt)