Zenith-9B-CodeCore-Merge-MLX

Zenith-9B-CodeCore-Merge is a merged 9B-parameter coding and reasoning model designed for long-horizon coding tasks, agentic coding, and agentic reasoning. It is built by merging Qwen3.5-9B as the base model with OxCoder-9B, Qwopus3.5-9B-Coder, and Ornith-1.5-9B, combining their capabilities for code generation, multi-step problem solving, instruction following, and autonomous coding workflows. The model is intended for complex software-engineering tasks that require sustained reasoning across multiple steps, code understanding, modification, debugging, and tool-oriented agentic workflows. This model is experimental and may generate artifacts.

• GGUF: Zenith-9B-CodeCore-Merge-GGUF. Note: The Multi-Token Prediction (MTP) heads are not preserved in this format. The model runs as a standard single-token-per-step autoregressive decoder.

Repository layout

+-- prithivMLmods/Zenith-9B-CodeCore-Merge-MLX (main)
    +-- / (Root: bf16)
    +-- 4bit/ (Quantized: 4-bit)
    +-- 8bit/ (Quantized: 8-bit)

Use with mlx

Install the required library:

pip install -U mlx-vlm

Model Note: Zenith-9B-CodeCore-Merge is a 9-billion parameter multimodal coding model designed for code generation, visual debugging, repository reasoning, and architecture diagram analysis. It supports both text and image/screenshot inputs.

BF16 Variant (Base Weights)

The full-precision BF16 files reside directly in the root of the repository:

CLI (Terminal)

python -m mlx_vlm generate \
  --model prithivMLmods/Zenith-9B-CodeCore-Merge-MLX \
  --max-tokens 512 \
  --temperature 0.0 \
  --prompt "Analyze this code screenshot, explain what it does, and fix any potential bugs or edge cases." \
  --image <path_to_image>

Python API

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

model_path = "prithivMLmods/Zenith-9B-CodeCore-Merge-MLX"
model, processor = load(model_path)
config = load_config(model_path)

image = ["<path_to_image>"]
prompt = "Analyze this code screenshot, explain what it does, and fix any potential bugs or edge cases."
formatted_prompt = apply_chat_template(processor, config, prompt, num_images=len(image))

output = generate(
    model, 
    processor, 
    formatted_prompt, 
    image=image, 
    max_tokens=512, 
    temperature=0.0
)
print(output.text)

8-bit Variant

Target the 8bit subfolder:

CLI (Terminal)

python -m mlx_vlm generate \
  --model prithivMLmods/Zenith-9B-CodeCore-Merge-MLX/8bit \
  --max-tokens 512 \
  --temperature 0.0 \
  --prompt "Analyze this code screenshot, explain what it does, and fix any potential bugs or edge cases." \
  --image <path_to_image>

Python API

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

model_path = "prithivMLmods/Zenith-9B-CodeCore-Merge-MLX"
model, processor = load(model_path, subfolder="8bit")
config = load_config(model_path, subfolder="8bit")

image = ["<path_to_image>"]
prompt = "Analyze this code screenshot, explain what it does, and fix any potential bugs or edge cases."
formatted_prompt = apply_chat_template(processor, config, prompt, num_images=len(image))

output = generate(
    model, 
    processor, 
    formatted_prompt, 
    image=image, 
    max_tokens=512, 
    temperature=0.0
)
print(output.text)

4-bit Variant

Target the 4bit subfolder:

CLI (Terminal)

python -m mlx_vlm generate \
  --model prithivMLmods/Zenith-9B-CodeCore-Merge-MLX/4bit \
  --max-tokens 512 \
  --temperature 0.0 \
  --prompt "Analyze this code screenshot, explain what it does, and fix any potential bugs or edge cases." \
  --image <path_to_image>

Python API

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

model_path = "prithivMLmods/Zenith-9B-CodeCore-Merge-MLX"
model, processor = load(model_path, subfolder="4bit")
config = load_config(model_path, subfolder="4bit")

image = ["<path_to_image>"]
prompt = "Analyze this code screenshot, explain what it does, and fix any potential bugs or edge cases."
formatted_prompt = apply_chat_template(processor, config, prompt, num_images=len(image))

output = generate(
    model, 
    processor, 
    formatted_prompt, 
    image=image, 
    max_tokens=512, 
    temperature=0.0
)
print(output.text)

License and Attribution

This model is based on and/or incorporates the following open-source projects and models:

• Qwen3.5-9B (Base): https://huggingface.co/Qwen/Qwen3.5-9B
• Zenith-9B-CodeCore-Merge: https://huggingface.co/prithivMLmods/Zenith-9B-CodeCore-Merge
• mlx-vlm: https://github.com/Blaizzy/mlx-vlm
• MLX: https://github.com/ml-explore/mlx

This model is released under the Apache License 2.0.