Midjourney Video Economics: Fast GPU Minutes, Batch Size and Relax Eligibility

Plan price and fast-minute entitlement

Midjourney’s plan guide identifies Basic at $10 monthly, Standard $30, Pro $60 and Mega $120. Fast GPU allowances are 200 minutes, 15 hours, 30 hours and 60 hours per month respectively. It also shows annual subscriptions as a separate commitment; $8 a month on Basic means $96 paid for the year, not a month-to-month checkout price. These web plan charges should remain separate from consumed GPU minutes.

Video costs are billed in GPU time

The first-party video guide lists an SD batch of four five-second videos at eight GPU minutes, two at four minutes and one at two. HD batches cost 26, 13 or seven GPU minutes for four, two or one output respectively. A small batch can therefore conserve allowance when only one candidate is required. The units are approximate GPU processing minutes, not generated wall-clock minutes or credits transferable to another vendor.

Relax and privacy have tier gates

Only Pro and Mega include video Relax Mode for SD resolution in the compared plans; Standard includes image Relax but not the same video entitlement. HD video requires a qualifying paid tier and Fast Mode. Stealth Mode for private images and videos is also restricted to Pro and Mega. A cost comparison without privacy and delivery requirements can incorrectly portray a cheaper tier as equivalent.

Costing iterations rather than outputs

If a studio makes twenty HD prompts as single-video batches, the documented sample uses approximately twenty times seven, or 140 Fast GPU minutes. Choosing four-output HD batches would instead use 520 Fast minutes for twenty prompts, but would generate many more candidates. Whether that is efficient depends on the acceptance rate. GPU consumption can vary by settings, and the table describes reference costs rather than an audited service-level commitment.

Fast time resets and additional GPU purchase

The provider’s plan documentation lists purchasing extra GPU time at $4 per hour. Its GPU guide says unused monthly Fast time does not roll over. Distinguish included allowance, paid additional hours, overage eligibility and slow Relax activity in records. Future model versions may have different costs, while video source-media rights and export licensing are governed by a different set of rules than the price schedule.

How to verify the actual invoice

Begin with a small representative workload rather than the cheapest published headline. Record the exact provider and billing channel, API model identifier, currency, account geography, plan and commitment length. The price or quota evidence in this guide is keyed to additional_gpu_hour, fast_gpu_minutes_per_4x5s_video, monthly_fast_gpu_minutes, monthly_subscription. Do not combine proprietary creative credits, token meters or media seconds into a single currency or invent a conversion rate between them. If an enterprise or regional rate is not published, leave its price unknown and obtain a written quote.

Keep separate estimates for failed attempts, optional tools, storage, exports, human quality review and tax. Use source-verified line items from the companion CSV instead of typing permanent numbers into your forecasting formulas. Where an old plan or model was superseded, preserve its original historical record and create an additive later snapshot. This guide does not claim a product test, performance ranking, independent legal assessment or live procurement approval.

Availability, alternatives and editorial limits

Available plans, supported endpoints, trial eligibility, quota reset and contract rights may vary by region and account even when a public website displays a single global currency. Compare an alternative only with the same accepted task, input/output volume, model version and privacy requirements. Vendor-provided marketing claims are not independent evidence of quality. This historical pricing article may be cross-referenced by the model, media or enterprise procurement team, but its purpose is to explain the billing mechanics without duplicating their evaluations or inventing test scores.