AI Image Generation for Esoteric Content: A Practical Workflow

The crystal ball stock photos are overused. Everyone in the tarot space has seen the same deck of Rider-Waite-adjacent images on ten different practitioners' websites. If your brand visuals look like everyone else's, potential clients have no reason to remember you specifically.

AI image generation solves this - but the workflow matters as much as the tool. A practitioner who generates 50 random images and picks the best one will get inconsistent results that don't form a recognizable visual identity. A practitioner who builds a repeatable prompt system produces a consistent aesthetic that becomes part of their brand.

This guide covers the practical setup: which tools, why, how to prompt them for esoteric content specifically, and how to combine AI generation with design tools to get production-ready assets.

The Two-Tool Setup That Covers Most Needs

Two tools cover roughly 90% of what a spiritual business needs for visual content:

Ideogram Plus ($15/month): text-in-images at high accuracy. Tarot card titles that are actually readable. Poster headlines. Course cover text. Promotional materials where the words matter. Ideogram v3 renders text at roughly 95% accuracy - put the words you want in quotes in your prompt and they appear where you asked.

Midjourney Standard ($30/month, or $24/month annual): artistic quality for images without embedded text. The atmospheric celestial fields, the painterly ritual scenes, the high-end mystical portrait aesthetics. Midjourney v7 produces the best raw artistic output of any major generator for the occult/mystical/fantasy aesthetic space.

If budget is tight, start with Ideogram Plus alone at $15/month. It handles text-heavy designs (which are most of what practitioners actually need - course covers, event posters, social graphics with quotes) and produces good artistic results even without text. You can always add Midjourney later for purely artistic hero images.

Adobe Firefly ($9.99/month standalone) matters specifically if you need commercial indemnification - Adobe's formal guarantee that they back you on copyright claims. For paid advertising, merchandise like printed tarot decks, or any high-stakes commercial use, Firefly is the safer legal ground. For social media and course materials, the indemnification matters less.

Source: docs.midjourney.com (official, 2026); eesel.ai/blog/ideogram-pricing (2026); aibusinessweekly.net (2026)

The Commercial Rights Issue

This comes up constantly and it's worth being direct about:

Midjourney Basic at $10/month does not include commercial use rights. If you use Basic-generated images in paid products, paid advertising, or revenue-generating content, you're violating Midjourney's terms. You need Standard ($30/month) minimum for commercial work.

Ideogram Free tier makes all your generated images public - visible to the entire Ideogram community. For client-facing branded materials, this is a privacy and originality problem. Ideogram Plus ($15/month) gives you private generations with commercial use.

US Copyright Office 2023 ruling: AI-generated images without substantial human authorship cannot be copyrighted by the prompter. This means similar prompts from different users produce legally usable images that no one "owns." Your competitive protection comes from using AI generation as a starting point and adding unique branded elements - not from claiming exclusivity over the generated image itself.

Prompting for Mystical Imagery

Vague prompts produce generic results. The aesthetic markers that reliably unlock good esoteric imagery:

For Midjourney:
- Style foundations: "art nouveau," "dark academia," "celestial," "sacred geometry," "cottagecore witchy"
- Technical quality markers: "highly detailed," "4k," "dramatic lighting," "painterly"
- Example: `mystical tarot card, The Moon card, sacred geometry, deep navy and silver, art nouveau, highly detailed, 4k`
- NOT: `tarot card moon` (too vague, gets generic results)

For Ideogram:
- Put text in quotes inside the prompt: `spiritual poster with text "Full Moon in Scorpio", watercolor, deep violet and gold tones`
- Be specific about layout: `centered text, wide margins, decorative border`
- Text in Ideogram appears where you describe it - describe the layout you want

For Adobe Firefly:
- Period-specific art styles work well: `1920s occult illustration style,` `medieval manuscript botanical,` `art deco mystical`
- Firefly's training on licensed stock content means it responds well to style descriptions that reference recognizable historical aesthetics

Step-by-Step Production Workflow

This is the process that produces consistent, usable assets - not a single lucky generation.

Step 1: Generate options Produce 4-8 variations using your prompt. Both Midjourney and Ideogram show multiple results per generation. Pick the strongest 1-2 candidates, not the first thing that appears.

Step 2: Enhance the selected image in Canva Pro The AI-generated image is raw material. In Canva Pro, you add: your logo, your brand colors on text elements, your specific typography, any additional text that needs to be in your exact brand font. This is where the image becomes yours rather than a prompt output.

For Midjourney images that need text added: generate without text in Midjourney (better artistic quality that way), then add all text in Canva Pro using your brand fonts.

Step 3: Magic Resize Canva Pro's Magic Resize converts one source design into all your required formats: Instagram Post (1:1), Instagram Story (9:16), Pinterest (2:3), LinkedIn (1.91:1), Facebook banner. One image, one resize action, all your channels.

Step 4: Apply Brand Kit If you've set up a Brand Kit in Canva Pro with your color palette and fonts, new designs automatically pull from those settings. See the Canva Brand Kit setup guide for the full configuration process.

Step 5: Save your prompts This is the step most people skip and regret. Keep a document with your best-performing prompts. Include the seed number from Midjourney (visible in the generation output) if you want to be able to reproduce a specific style. Six months later when you need more images in the same aesthetic, you'll thank yourself.

Managing Brand Consistency

AI generation has a consistency problem: the same prompt produces similar but not identical results. Two images from the same prompt look related but not like they came from the same photograph session.

For a cohesive visual brand, you need one or more of:
- A fixed color palette applied in Canva on top of every generation (your brand colors, not whatever the AI chose)
- A signature compositional style in your prompts (always "centered subject, dark background," or always "wide landscape with figure in foreground")
- Consistent application of your logo and brand elements as an overlay in Canva

The AI provides the raw aesthetic. Canva provides the brand consistency. Neither alone does both.

For the full comparison of these three AI tools side by side, see Midjourney vs Ideogram vs Adobe Firefly for esoteric content. For how to build a broader content calendar around the images you create, see batch content creation for spiritual practitioners. For Canva compared to other design tools, see Canva vs Adobe Express.

Cost Summary

Budget setup (one tool): Ideogram Plus $15/month + Canva Pro $15/month = $30/month total.

Full setup: Ideogram Plus + Midjourney Standard + Canva Pro = $60-69/month. Covers every image type a spiritual business needs at professional quality.

Frequently Asked Questions

Do I need both Ideogram and Midjourney or can I use just one?

Start with one. Ideogram Plus at $15/month handles text-heavy designs (which is most content: course covers, posters, social graphics with quotes or headlines). If you also need purely artistic images without text - atmospheric hero images for your website, for example - add Midjourney Standard later. Many practitioners find Ideogram alone covers 80% of their needs.

Can I sell products with AI-generated images on them?

Yes, with the right plan. Midjourney Standard ($30/month) and Ideogram Plus ($15/month) both include commercial use rights. What you cannot copyright is the AI generation itself - so a competitor could theoretically use a similar prompt and get a similar image. Add your own design elements in Canva to create derivative work that's more distinctly yours.

What if the AI generates something that looks like an existing tarot deck?

Rider-Waite-Smith's original illustrations are in the public domain. However, specific modern editions and published reproductions may have their own copyright. If you're generating imagery that closely resembles a specific contemporary published deck, be cautious. AI-generated imagery inspired by general tarot symbolism and medieval iconography (rather than specific copyrighted artwork) is significantly safer ground.

How do I keep my prompts consistent across different sessions?

Save every prompt you want to reuse in a dedicated document - a Notion page or Google Doc works fine. Include the Midjourney seed number if you want to reproduce a specific variation. When returning to a project weeks later, the saved prompt is the fastest path back to the same aesthetic.

Is it worth using Adobe Firefly if I already have Midjourney?

For most solo practitioners: only if you need the commercial indemnification for a specific high-stakes project (merchandise, large-scale paid advertising). For day-to-day social media and course content, Midjourney Standard's commercial rights are sufficient. The indemnification matters most when the financial exposure from a copyright claim would be significant.