Midjourney does not bill by raw image counts. Instead, it meters high-priority GPU compute time, workflow throughput, and commercial asset privacy. Choosing between the Basic Plan ($10/month), Standard Plan ($30/month), and Pro Plan ($60/month) comes down to three operational thresholds:
- The Math of GPU Depletion: If you consume more than 5 additional Fast GPU hours in a single month on Basic, you have already spent enough to pay for Standard ($10 base + $20 top-up = $30). If you buy more than 7.5 extra Fast hours on Standard, you are overspending compared to upgrading directly to Pro ($30 base + $30 top-up = $60).
- Creative Exploration vs. Meter Anxiety: Basic gives you no safety net. Once its 3.3 Fast hours run out, generation stops completely unless you purchase add-on credits. Standard and Pro unlock unlimited Relax Mode, allowing artists to iterate, test prompts, and explore compositions without burning billable compute.
- Client Confidentiality and NDA Compliance: Direct messaging the Midjourney bot on Discord does not keep your work private. By default, every generation appears on the public Midjourney community gallery. Stealth Mode—the only mechanism to hide prompts and outputs from public web indices—is strictly exclusive to the Pro Plan ($60/month) and above.
Here is the exact technical breakdown, operational math, and workflow criteria to determine which subscription tier genuinely justifies your investment.
Subscription Tier Breakdown at a Glance
The following matrix compares the core technical entitlements across Midjourney's primary subscription plans. Annual billing provides a 20% discount across all tiers.
| Feature / Quota | Basic Plan | Standard Plan | Pro Plan | Mega Plan (High-Volume) |
|---|---|---|---|---|
| Monthly Price (Annual Rate) | $10/mo ($8/mo / $96/yr) | $30/mo ($24/mo / $288/yr) | $60/mo ($48/mo / $576/yr) | $120/mo ($96/mo / $1,152/yr) |
| Fast GPU Time | 3.3 hours / month (~200 min) | 15 hours / month | 30 hours / month | 60 hours / month |
| Est. Fast Grid Outputs | ~200 generations (4-image grids) | ~900–1,000 generations | ~1,800–2,000 generations | ~3,600–4,000 generations |
| Relax Mode (Unlimited) | Not available | Included (Images & SD Video) | Included (Images & SD Video) | Included (Images & SD Video) |
| Stealth Mode (Private Gallery) | Not available (Public) | Not available (Public) | Included (/stealth) | Included (/stealth) |
| Max Concurrent Fast Jobs | 3 active jobs | 3 active jobs | 12 active jobs | 12 active jobs |
| Max Concurrent Relax Jobs | 0 (No Relax Mode) | 3 active jobs | 3 active jobs | 3 active jobs |
| Max Queued Tasks | 10 in queue | 10 in queue | 10 in queue | 10 in queue |
| Permutation Prompt Limit | Up to 4 variations | Up to 10 variations | Up to 40 variations | Up to 40 variations |
| Extra Fast GPU Cost | $4 per hour | $4 per hour | $4 per hour | $4 per hour |
| Commercial Usage Rights | Included (Under $1M gross rev) | Included (Under $1M gross rev) | Mandatory for >$1M gross rev | Mandatory for >$1M gross rev |
1. Fast GPU Hours & The Top-Up Break-Even Math
Midjourney measures standard image generation in GPU compute seconds. A standard prompt generating a four-image grid typically takes between 30 and 50 seconds of GPU time (averaging ~0.7 to 0.8 minutes). Upscaling, panning, zooming, and running high-detail parameters increase compute consumption accordingly.
Unused Fast GPU hours do not roll over across billing cycles; they reset on your monthly renewal date.
The Basic Plan Limit ($10/mo, 3.3 Fast Hours)
The Basic Plan grants 3.3 hours of Fast GPU compute (approximately 200 minutes). In practical creative terms:
- That equals roughly 6 to 7 prompt generations per calendar day.
- If you run multiple variations, re-rolls, and upscales, you can burn through this entire monthly allowance in a single afternoon brainstorming session.
- Once this quota is exhausted, Midjourney refuses all new
/imaginecommands until you either wait for the next billing cycle or purchase add-on Fast GPU time at $4 per hour.
The Financial Tipping Points
Midjourney allows subscribers to purchase supplemental Fast GPU hours at a flat rate of $4 per hour. However, relying on a la carte hours frequently leads to bad unit economics:
- Basic to Standard Break-Even: Upgrading from Basic ($10) to Standard ($30) is a $20 difference. At $4 per hour, $20 buys precisely 5 hours of add-on GPU time. If your creative workflow requires more than 8.3 total hours of Fast GPU time in a month, staying on Basic and buying add-on credits is mathematically worse than subscribing to Standard—and you miss out on unlimited Relax Mode entirely.
- Standard to Pro Break-Even: Upgrading from Standard ($30) to Pro ($60) represents a $30 difference. That $30 buys 7.5 hours of add-on compute. If you find yourself consistently buying 8 or more extra hours on Standard to meet client deadlines, switching to Pro gives you double the baseline Fast GPU quota (30 hours total) while unlocking Stealth Mode and a 4x increase in concurrent generation speed.
2. Relax Mode Access: Eliminating the Cost of Creative Iteration
The defining functional boundary between Basic and Standard is Relax Mode.
In Fast Mode, jobs immediately dispatch to high-priority GPU clusters. In Relax Mode, jobs enter a dynamic fair-use queue. While a Fast generation returns in 30 to 50 seconds, a Relax generation may take anywhere from 1 to 10 minutes depending on community traffic and your recent utilization.
Why Relax Mode Transforms the Creative Process
Concept artists, illustrators, and visual directors do not know their final prompt at the start of a project. Professional concept development requires exploring camera angles, color palettes, lighting keys, and seed combinations.
- On Basic, every failed prompt, typo, or minor seed change deducts roughly 45 seconds of your non-refundable $10 monthly budget. This creates friction and conservative, risk-averse prompting.
- On Standard and Pro, you can run your early ideation in Relax Mode (
/relax). You can fire off hundreds of wide-angle character variations, style explorations, and aesthetic tests without depleting a single second of your Fast hour quota. - When client feedback arrives or a tight deadline demands instantaneous results, you toggle back to Fast Mode (
/fast) to generate final compositions, select high-resolution crops, and execute fine variations.
For any creator working on commercial client briefs or personal IP, Relax Mode pays for the $20 upgrade difference between Basic and Standard by removing the mental penalty on experimental drafting.
3. Stealth Mode: The Privacy Prerequisite for Commercial Work
A widespread misunderstanding among Midjourney users is that generating images within Discord Direct Messages (DMs) or on a private Discord server keeps assets confidential. It does not.
The Default Public Gallery
Midjourney is architected as an open community. Regardless of whether an image is generated in a shared public Discord channel or inside your private direct message thread with the Midjourney Bot, the resulting image, prompt text, parameter string, and seed number are automatically uploaded and indexed on the public Midjourney community showcase (midjourney.com/explore).
Any user or competitor can:
- Search your Midjourney profile and view your entire production history.
- Search keywords and view unreleased concept art, character sheets, and marketing mockups.
- Copy your exact prompt strings, weights, and image references (
--sref,--cref).
How Stealth Mode Works
Stealth Mode (/stealth) is restricted strictly to the Pro ($60/month) and Mega ($120/month) tiers.
When Stealth Mode is active:
- Generations do not appear in the public Midjourney Explore feed or search index.
- Your creations remain visible only to you within your personal gallery on the Midjourney web platform.
- You can selectively make specific images public using the
--publicparameter if desired.
Who Needs Stealth Mode?
If your work touches any of the following categories, the upgrade to Pro is not optional—it is a mandatory legal and operational safeguard:
- Commercial Designers Under Client NDAs: Producing concept art for unannounced video games, films, advertising campaigns, or industrial design requires complete secrecy. Uploading client-specific IP to a public index can constitute a direct breach of non-disclosure agreements.
- Proprietary Prompt Engineering: Studios and agency teams developing distinctive brand styles or proprietary prompt formulas cannot risk leaking their custom syntax and parameter chains to public scrapers.
- Competitive Brand Strategy: Agencies preparing pitch decks or seasonal packaging mockups risk exposing campaign ideas before official launch.
4. Concurrency & Permutations: Production Throughput Comparison
When managing high-volume production schedules, generation queuing becomes a major bottleneck. The Pro Plan fundamentally alters generation throughput compared to Basic and Standard.
Concurrent Job Capacities
- Basic & Standard: Limited to 3 concurrent active Fast jobs at any given moment. Any additional requests are placed into a backlog queue (up to 10 waiting jobs).
- Pro & Mega: Expanded to 12 concurrent active Fast jobs simultaneously, alongside 3 concurrent Relax jobs and 10 queued requests.
Running 12 concurrent threads allows a visual designer to produce 48 individual visual concepts (12 grids of 4) in parallel under one minute, compared to waiting through four sequential rounds on lower tiers.
Permutation Prompts
Midjourney's permutation syntax allows users to test multiple variables in a single prompt line using curly brackets:
/imagine prompt cinematic still of a cybernetic warrior in {rain, snow, sandstorm} at {dawn, noon, midnight} --ar 16:9
This single prompt expands into 9 distinct generation tasks (3 weather conditions multiplied by 3 times of day).
- Basic Plan: Capped at a maximum of 4 variations per permutation prompt.
- Standard Plan: Capped at a maximum of 10 variations per permutation prompt.
- Pro Plan: Capped at up to 40 variations per permutation prompt.
For commercial look-development and style bracketing, the ability to launch 40 variations with a single prompt submission transforms a multi-hour manual task into an automated single-click operation.
5. Enterprise Legal Terms: The $1,000,000 Revenue Threshold
Midjourney’s official Terms of Service include a specific legal boundary governing commercial asset ownership.
While subscribers across all paid plans (Basic, Standard, Pro, and Mega) receive commercial usage rights for the assets they create, Section 4 of Midjourney's Terms establishes an exception:
- Companies and entities generating more than $1,000,000 USD in gross annual revenue (including employees and contractors working on their behalf) are legally required to maintain an active Pro or Mega subscription in order to own and commercially exploit their generated assets.
- If you work for a mid-to-large agency, enterprise brand, or venture-backed studio with revenues exceeding this threshold, you cannot legally deploy Midjourney assets commercially under Basic or Standard plans.
Decision Framework: Which Tier Should You Choose?
Choose the Basic Plan ($10/month) if:
- You are a beginner exploring AI image generation and learning prompt fundamentals.
- You produce fewer than 10–15 prompts per week and do not rely on Midjourney for daily creative deliverables.
- Your work is non-commercial or personal, meaning public visibility on the community showcase poses no intellectual property concern.
Upgrade to the Standard Plan ($30/month) if:
- You are a practicing freelance illustrator, independent creator, or digital artist who generates images regularly throughout the week.
- You want the freedom of unlimited Relax Mode to explore creative ideas, refine composition, and test seeds without monitoring remaining compute hours.
- Your project work is public, personal, or does not require strict non-disclosure protections.
- You typically consume between 10 and 25 hours of total GPU time per month.
Upgrade to the Pro Plan ($60/month) if:
- You work with commercial clients under NDAs, handle proprietary intellectual property, or require absolute image privacy via Stealth Mode.
- You run high-volume production cycles requiring 12 concurrent Fast jobs and extensive permutation tests (up to 40 variations).
- You routinely exhaust Standard's 15 Fast hours and spend $30 or more per month buying a la carte compute credits ($4/hour).
- Your employer or corporate entity exceeds $1,000,000 USD in annual gross revenue.