Instructions to use zimamedia/zenith-v0.1-f16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zimamedia/zenith-v0.1-f16 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="zimamedia/zenith-v0.1-f16")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("zimamedia/zenith-v0.1-f16")
model = AutoModelForImageTextToText.from_pretrained("zimamedia/zenith-v0.1-f16")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use zimamedia/zenith-v0.1-f16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zimamedia/zenith-v0.1-f16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zimamedia/zenith-v0.1-f16",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/zimamedia/zenith-v0.1-f16

SGLang

How to use zimamedia/zenith-v0.1-f16 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "zimamedia/zenith-v0.1-f16" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zimamedia/zenith-v0.1-f16",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "zimamedia/zenith-v0.1-f16" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zimamedia/zenith-v0.1-f16",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Unsloth Studio new

How to use zimamedia/zenith-v0.1-f16 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for zimamedia/zenith-v0.1-f16 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for zimamedia/zenith-v0.1-f16 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for zimamedia/zenith-v0.1-f16 to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="zimamedia/zenith-v0.1-f16",
    max_seq_length=2048,
)

Docker Model Runner
How to use zimamedia/zenith-v0.1-f16 with Docker Model Runner:
```
docker model run hf.co/zimamedia/zenith-v0.1-f16
```

Zenith v0.1 — CMO Ready Agent by Zima Media

Developed by: Zima Media
License: apache-2.0
Finetuned from model: unsloth/gemma-4-E2B-it

What is Zenith?

Zenith v0.1 is a CMO Ready Agent powered by the expertise of Mike Zima, Chief Marketing Officer at Zima Media.

This model captures Mike's deep knowledge across:

Marketing strategy & planning — go-to-market, brand positioning, campaign architecture
Sales prospecting & lead generation — B2B outreach, CRM workflows, data-driven acquisition
Content marketing & storytelling — narrative frameworks, thought leadership, audience engagement
Agency operations & client management — scaling services, team leadership, client success

Zenith is designed as a private AI tool for CMOs and growth leaders to safely test marketing ideas, strategy frameworks, and messaging without exposing sensitive data or relying on third-party models.

Why This Matters

Marketing is not broken. Your system is.

Most agencies sell channel execution. They don't build the systems that drive revenue. Zenith is built on a different philosophy — one system, one strategy, one outcome: profitable growth. No silos. No fragmentation. No channel-first thinking.

This is a CMO Ready Agent for leaders who want to:

Test new marketing ideas privately before committing budget
Validate strategy frameworks with an AI trained on real-world expertise
Move faster without exposing sensitive data to external APIs
Think bigger with a thinking partner that understands growth at scale

Training Data Breakdown

Zenith v0.1 was trained on a highly curated selection representing just 1% of our total available data — carefully chosen to capture the highest-signal content across Mike Zima's professional expertise. The remaining 99% was excluded to ensure quality, relevance, and a focused knowledge base.

The training corpus draws from the following sources:

Source	Percentage of Training Data
LinkedIn posts & articles	30%
Blog content & website copy	25%
Internal strategy documents & playbooks	20%
Presentations & slide decks	15%
Video content (recordings, talks)	10%

Expertise Areas Covered

The model was fine-tuned on Mike Zima's deep knowledge across:

Marketing strategy & planning — go-to-market frameworks, brand positioning, campaign architecture, budget allocation
Sales prospecting & lead generation — B2B outreach strategies, CRM workflows, data-driven acquisition, list building
Content marketing & storytelling — narrative frameworks, thought leadership, audience engagement, SEO content strategy
Agency operations & client management — scaling services, team leadership, client success frameworks, pricing models
Growth marketing & performance — paid media strategy, conversion optimization, analytics-driven decision making
Brand development — brand identity, messaging architecture, competitive differentiation

DIY Installation Guide (macOS & Linux)

Getting Zenith up and running locally is straightforward. Here's how to install it for the first time:

Prerequisites

Python 3.10+
Hugging Face account (free) and a valid access token
At least 16GB RAM (32GB recommended for smooth inference)

Step 1: Install Unsloth (Recommended)

The easiest way to get started is with the official pip package. It pulls in all required dependencies automatically:

pip install unsloth

For GPU support (NVIDIA), Unsloth will auto-detect and configure CUDA. For Apple Silicon (M1/M2/M3), it uses the native Metal backend — no extra setup needed.

Step 2: Authenticate with Hugging Face

huggingface-cli login

Step 3: Load & Run the Model

Use this quick Python script to load Zenith locally:

from unsloth import FastLanguageModel
import torch

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name="zimamedia/zenith-v0.1", # or your hosted repo
    max_seq_length=2048,
    load_in_4bit=True,
)

FastLanguageModel.for_inference(model)

messages = [
    {"role": "user", "content": "Write a go-to-market strategy for a SaaS product."}
]

input_ids = tokenizer.apply_chat_template(messages, return_tensors="pt").to(model.device)
output = model.generate(input_ids, max_new_tokens=512, use_cache=True)
print(tokenizer.decode(output[0], skip_special_tokens=True))