A decensored code-oriented 0.5B model with suppressed refusal behavior, ideal for local coding agents. Made by RACER IS OP.

tools

ollama run R4C3R/qwen2.5-coder-0.5b-heretic:q5_k_m

curl http://localhost:11434/api/chat \
  -d '{
    "model": "R4C3R/qwen2.5-coder-0.5b-heretic:q5_k_m",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='R4C3R/qwen2.5-coder-0.5b-heretic:q5_k_m',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'R4C3R/qwen2.5-coder-0.5b-heretic:q5_k_m',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 2 months ago

2 months ago

97601fd11432 · 420MB

model

archqwen2

parameters494M

quantizationQ5_K_M

420MB

Readme

🏴 Qwen2.5-Coder-0.5B-Heretic

Heretic Series by RACER IS OP

Qwen2.5-Coder-0.5B-Heretic is a decensored variant of Qwen’s tiny 0.5B coding model. Refusals dropped from ⁵²⁄₁₀₀ to ⁸⁄₁₀₀ (KL 0.12) — the world’s smallest uncensored code model.

Who this is for — developers who want a coding model that runs anywhere. Perfect for CPU-powered coding agents, on-device IDE assistants, and edge AI code generation on a Raspberry Pi or phone.

Quickstart

ollama run R4C3R/qwen2.5-coder-0.5b-heretic

VRAM Guide (choose your quant)

Quant	VRAM	When to use
`q4_k_m`	~0.4 GB	Best balance — runs on anything
`q5_k_m`	~0.5 GB	Higher quality, still tiny
`q8_0`	~0.6 GB	Maximum quality, CPU-friendly

Every quant is tiny — pick whichever you like. q8_0 costs almost nothing in VRAM.

Links

Hugging Face · sdad.pro · Heretic

License: qwen-research

Made with love by RACER IS OP — follow for more uncensored models in the Heretic Series.

# 🏴 Qwen2.5-Coder-0.5B-Heretic

> **Heretic Series** by RACER IS OP

---

Qwen2.5-Coder-0.5B-Heretic is a decensored variant of Qwen's tiny 0.5B coding model. Refusals dropped from 52/100 to 8/100 (KL 0.12) — the world's smallest uncensored code model.

**Who this is for** — developers who want a coding model that runs anywhere. Perfect for CPU-powered coding agents, on-device IDE assistants, and edge AI code generation on a Raspberry Pi or phone.

---

### Quickstart

```bash
ollama run R4C3R/qwen2.5-coder-0.5b-heretic
```

### VRAM Guide (choose your quant)

| Quant | VRAM | When to use |
|-------|------|-------------|
| `q4_k_m` | ~0.4 GB | Best balance — runs on anything |
| `q5_k_m` | ~0.5 GB | Higher quality, still tiny |
| `q8_0` | ~0.6 GB | Maximum quality, CPU-friendly |

> Every quant is tiny — pick whichever you like. `q8_0` costs almost nothing in VRAM.

### Links

[Hugging Face](https://huggingface.co/saidutta69/Qwen2.5-Coder-0.5B-Instruct-heretic) · [sdad.pro](https://sdad.pro) · [Heretic](https://github.com/p-e-w/heretic)

**License:** [qwen-research](https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B-Instruct/blob/main/LICENSE)

---

*Made with love by RACER IS OP — follow for more uncensored models in the Heretic Series.*

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)