---
license: ecl-2.0
base_model:
- openai/gpt-oss-20b
tags:
- Claude
- Fable
- Antropic
- Agent
- Ollama
- vLLM
- mxfp4
- quantization
- moe
widget:
- text: "-"
  output:
    url: images/Claude-OSS-Fable-5.webp
datasets:
- Tesleum/Fable-5-traces-Harmony
pipeline_tag: text-generation
library_name: transformers

# This structures your data into Hugging Face's official widget database
model-index:
- name: 0xCoder
  results:
  - task:
      type: text-generation
    dataset:
      name: Model Specifications
      type: custom
    metrics:
    - name: True Architecture Size
      type: parameters
      value: "21B Total Params (3.6B Active)"
    - name: Native Hardware Format
      type: format
      value: "MXFP4 (Microscaling 4-bit)"
---

# 📝 0xCoder (Fable 5)

> [!NOTE]
> **Hugging Face Metadata:** The automated metadata tag at the top of this page miscalculates this model as a 12B BF16/U8 model because the Hub parser cannot natively calculate the 32-expert MoE layout of `Claude-OSS`. The actual model size is **21B parameters** running natively on **MXFP4 microquantization**.

<Gallery />

## Model description 

Open-Source OpenAI model powered by Claude Fable 5 Agent

# Highlights

* **Claude Fable 5:** Fully Train model to Calude Fable 5 specific use case through parameter fine-tuning.
* **Agentic capabilities:** Use the models’ native capabilities for function calling, [web browsing](https://github.com/openai/gpt-oss/tree/main?tab=readme-ov-file#browser), [Python code execution](https://github.com/openai/gpt-oss/tree/main?tab=readme-ov-file#python), and Structured Outputs.
* **MXFP4 quantization:** The models were post-trained with MXFP4 quantization of the MoE weights, making model run within 16GB of memory. All evals were performed with the same MXFP4 quantization.
* **Configurable reasoning effort:** Easily adjust the reasoning effort (low, medium, high) based on your specific use case and latency needs.  
* **Full chain-of-thought:** Gain complete access to the model’s reasoning process, facilitating easier debugging and increased trust in outputs. It’s not intended to be shown to end users.  

---

# Inference examples

## Transformers

You can use `0xCoder` with Transformers. If you use the Transformers chat template, it will automatically apply the [harmony response format](https://github.com/openai/harmony). If you use `model.generate` directly, you need to apply the harmony format manually using the chat template or use our [openai-harmony](https://github.com/openai/harmony) package.

To get started, install the necessary dependencies to setup your environment:

```
pip install -U transformers kernels torch 
```

Once, setup you can proceed to run the model by running the snippet below:

```py
from transformers import pipeline
import torch

model_id = "Tesleum/0xCoder"

pipe = pipeline(
    "text-generation",
    model=model_id,
    torch_dtype="auto",
    device_map="auto",
)

messages = [
    {"role": "user", "content": "Explain quantum mechanics clearly and concisely."},
]

outputs = pipe(
    messages,
    max_new_tokens=256,
)
print(outputs[0]["generated_text"][-1])
```

Alternatively, you can run the model via [`Transformers Serve`](https://huggingface.co/docs/transformers/main/serving) to spin up a OpenAI-compatible webserver:

```
transformers serve
transformers chat localhost:8000 --model-name-or-path Tesleum/0xCoder
```

## vLLM

vLLM recommends using [uv](https://docs.astral.sh/uv/) for Python dependency management. You can use vLLM to spin up an OpenAI-compatible webserver. The following command will automatically download the model and start the server.

```bash
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Tesleum/0xCoder",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

vllm serve "Tesleum/0xCoder"
```

## Docker

```bash
docker model run hf.co/Tesleum/0xCoder
```
## Ollama

To achieve better performance and quality, use **vLLM** instead 

#### LM Studio

To achieve better performance and quality, use **vLLM** instead 

---

# Reasoning levels

You can adjust the reasoning level that suits your task across three levels:

* **Low:** Fast responses for general dialogue.  
* **Medium:** Balanced speed and detail.  
* **High:** Deep and detailed analysis.

The reasoning level can be set in the system prompts, e.g., "Reasoning: high".

# Tool use

The 0xCoder models are excellent for:
* Agentic operations like browser tasks
* Function calling with defined schemas
* Web browsing (using built-in browsing tools)