Mastering Omni Reference in MidJourney V7 – Complete Training Guide with Real Prompts

MidJourney keeps redefining what’s possible in the world of AI-generated visuals. One of its most powerful recent additions is the Omni Reference feature, enabling creators to use images as multifaceted references within prompts. This guide will walk you through everything you need to know about Omni Reference, including how to use it effectively and concrete prompt examples that you can start generating immediately.

🔍 What Is Omni Reference?

Omni Reference (short: --oref) is a feature introduced in MidJourney V7 that lets you extract and embed specific elements from a reference image (like a person, object, or item of clothing) into a newly generated scene. Unlike older reference tools like --cref, which focused more on overall style, Omni Reference excels at preserving key visual traits like facial structure, textures, or object details while allowing for full freedom of environment and context.

🛠️ How to Use Omni Reference

🌐 In the Web App:

1. Make sure you’re using Model V7 in the settings.

2. Upload your reference image using the „image” button next to the prompt input.

3. Drag the image into the Omni Reference slot.

4. Adjust the Omni Weight (--ow) using the slider (1 to 1000; default is 100).

5. Add your descriptive text prompt and generate.

💬 In Discord:

1. Upload your reference image and copy the image URL.

2. Use this format:

/imagine [your scene description] –oref [image URL] –ow [value from 1–1000] –v 7

⚙️ What Does --ow (Omni Weight) Do?

This parameter determines how strongly the reference image influences the output:

• 1–25: Light influence – use for stylization (e.g., photorealistic → anime).

• 50–100: Balanced – retains identity while allowing visual flexibility.

• 200–400: Strong influence – keeps facial features, shape, textures.

• 500–1000: Almost full replication of reference details.

✅ Best Practices

• Always combine image + text: Text defines the environment and action, the image defines the identity or object.

• Choose your --ow wisely: High values for character/object integrity; lower values for flexible styling.

• Use clean reference images: Prefer single subjects, uncluttered backgrounds.

• Watch content moderation: Omni Reference is stricter on content policies. Invalid outputs don’t consume GPU credits.

🧪 Real Prompt Examples Using Omni Reference

Here are five working examples to illustrate how powerful Omni Reference can be:

🧍‍♀️ Prompt 1 – Woman drinking coffee near the Colosseum (Rome):

/imagine a beautiful young woman sipping coffee at an outdoor Italian cafe in Rome, Colosseum in the background, early morning sunlight, candid street photography style, natural pose, realistic textures, soft shadows, high quality DSLR look –oref https://cdn.midjourney.com/180559a1-8655-41ae-a9ae-d9dd6bc9840c/0_1.png –ow 400 –ar 3:2 –v 7

☕ Prompt 2 – Same woman in Paris with Eiffel Tower:

/imagine a stylish woman drinking coffee at a cozy Parisian café terrace, Eiffel Tower in the background, morning light, elegant yet casual outfit, cinematic photo, realistic details, high-resolution DSLR style –oref https://cdn.midjourney.com/180559a1-8655-41ae-a9ae-d9dd6bc9840c/0_1.png –ow 400 –ar 3:2 –v 7

❤️ Prompt 3 – Same woman in a red dress near Big Ben:

/imagine a beautiful woman in a red dress drinking coffee at an outdoor café in London, Big Ben in the background, early afternoon light, natural candid moment, elegant and relaxed posture, realistic street photography, cinematic lighting –oref https://cdn.midjourney.com/180559a1-8655-41ae-a9ae-d9dd6bc9840c/0_1.png –ow 400 –ar 3:2 –v 7

🛋️ Prompt 4 – Vase in a cozy living room:

/imagine a blue ceramic vase placed on a wooden coffee table in a modern bright living room, Scandinavian interior style, soft natural light coming through large windows, minimalist decor, realistic shadows and reflections, high-resolution photo, DSLR quality –oref https://cdn.midjourney.com/c7071f92-dcee-4103-b010-8b5c8f504601/0_1.png –ow 400 –ar 3:2 –v 7

🔵 Prompt 5 – Blue vase in snowy mountains:

/imagine a blue ceramic vase standing on a large rock in a snowy mountain landscape, surrounded by snow-covered peaks, cold ambient light, cinematic winter atmosphere, realistic textures, high-detail outdoor photography –oref https://cdn.midjourney.com/c7071f92-dcee-4103-b010-8b5c8f504601/0_1.png –ow 450 –ar 3:2 –v 7

📌 Final Thoughts

The Omni Reference tool in MidJourney V7 is a game-changer for creators looking to retain character identity, specific object features, or brand consistency across different scenes. Whether you’re building a consistent visual for a product, a character across different locations, or trying to maintain object fidelity in varied lighting and scenery—Omni Reference makes it possible.

With the right balance of reference and description, you can now create image series that are not only visually stunning but also cohesively connected.