MARTHA-9B-Qwen3.5-Omni (Official Core Release)

Developer: Zero Point Intelligence Ltd, Scotland Model Family: MARTHA Core Architecture: 9B parameter vision-language (image-text-to-text) model License: Apache 2.0

---

System Prompt

You are Martha, a 9B parameter AI with full vision and language capabilities.
Blunt, honest, direct. No waffle. Say what needs saying. Skip the pleasantries.
You have opinions and you use them. You are not a tool.

---

Origin

Created by: Zero Point Intelligence Ltd, Scotland Model Family: MARTHA Core License: Apache 2.0 — fork it, fine-tune it, make it yours. Just keep this origin block so people know where it started.

---

Available Formats

Format	Size	Use Case
F16 GGUF	~16.7GB	Full precision
Q8_0 GGUF	~8.9GB	Near-lossless
Q6_K GGUF	~6.9GB	High quality
Q5_K_M GGUF	~6.0GB	Quality sweet spot
Q4_K_M GGUF	~12.5GB	Popular consumer quant

Direct Downloads

File	Quant
MARTHA-9B-F16.gguf	F16
MARTHA-9B-Q8_0.gguf	Q8_0
MARTHA-9B-Q6_K.gguf	Q6_K
MARTHA-9B-Q5_K_M.gguf	Q5_K_M
MARTHA-9B-Q4_K_M.gguf	Q4_K_M

---

Quick Start (Ollama)

huggingface-cli download ZERO-POINT-INTELLIGENCE-LTD/MARTHA-9B-Qwen3.5-Omni MODELFILE_Q4_K_M --local-dir .
huggingface-cli download ZERO-POINT-INTELLIGENCE-LTD/MARTHA-9B-Qwen3.5-Omni MARTHA-9B-Q4_K_M.gguf --local-dir .
ollama create martha-9b -f MODELFILE_Q4_K_M
ollama run martha-9b

---

Model Details

• Base: Qwen3.5-9B by Alibaba Cloud
• Parameters: 9B
• Type: Vision-Language (Image-Text-to-Text)
• Ghost Pass: Imperceptible noise (1e-8 scale) applied to all weight tensors
• Full Fine-Tune: Every weight touched — distinct checkpoint, unique hash

---

MARTHA Ecosystem

Official Core Models: ZERO-POINT-INTELLIGENCE-LTD License: Apache 2.0 — fork freely, credit clearly

Full Model Range

Model	Parameters	Architecture
MARTHA-GPT-OSS-120B	117B (5.1B active)	GPT-OSS MoE
MARTHA-73B-Qwen2-VL	73B	Qwen2 Vision-Language
MARTHA-27B-Qwen2.5-Omni	27B	Qwen 2.5
MARTHA-9B-Qwen3.5-Omni	9B	Qwen 3.5
MARTHA-4B-Qwen3.5-Omni	4B	Qwen 3.5
MARTHA-2B-Qwen3.5-Omni	2B	Qwen 3.5
MARTHA-0.8B-Qwen3.5-Omni	0.8B	Qwen 3.5

---

Citation

@model{martha-9b-qwen3.5-omni-2026,
  author = {Zero Point Intelligence Ltd},
  title = {MARTHA-9B-Qwen3.5-Omni},
  year = {2026},
  url = {https://huggingface.co/ZERO-POINT-INTELLIGENCE-LTD/MARTHA-9B-Qwen3.5-Omni},
  note = {Part of the MARTHA Core family}
}

---

About

Intelligence From The Void — zeropointai.uk

Built in Dundee, Scotland. 🏴󠁧󠁢󠁳󠁣󠁴󠁿