Plans & pricing

Trial (free for 1 month, no card)Starter $12/seat/moor $120/seat/yrPro $25/seat/moor $250/seat/yrMost popularBusiness $50/seat/moor $500/seat/yrSelf-Hosted †priced per deployment · Talk to us
Video generation time per month §10 minutes (watermarked export)60 minutes240 minutes600 minutes per seat, pooled across the teamYour hardware's capacity
Longest clip15 s15 s30 s90 sYour hardware's capacity, up to about 5 minutes
Images per month3006,00015,00045,000 per seat, pooledYour hardware's capacity
Resolution1080p1080p1080p1080p generation, 2K exportUp to 4K on suitable hardware
ProcessingEco: off-peak queueStandard: regular queueExpress: ahead of the regular queuePriority: first in lineRuns on customer hardware; no shared queue
Generations at once1248Set by the deployment
Storage5 GB50 GB200 GB500 GB per seat, pooledUp to 1 PB per deployment, on your hardware
Extra storage-$2.99 per 50 GB block a month, charged when you add it, renews with your plan-
Video top-up-+30 minutes for $4.99, one-time; unused minutes never expire while you are subscribed-
Pause-Pause your plan for up to 3 months; your files stay, nothing is charged-
Passwordless sign-in / SSO / 2FANo passwords, ever. Sign in with an account you already have (Google or Microsoft today, more coming), or with a 6-digit code sent to your email or mobile phone; nothing to remember, nothing for us to store. Your provider's two-step verification carries over. Company logins (Okta, Microsoft Entra, Google Workspace) work through the Meraqi Identity Broker (coming soon).The organization's own login system connects through the Meraqi Identity Broker over SAML 2.0 or OIDC (coming soon). Smart-card (CAC/PIV) sign-in is a paid add-on.

§ Generation time, what it means: the minutes are how long our GPUs work on your video, not the length of the finished clip. What a clip costs to generate depends on what you are making, the quality you ask for, how long it runs and which model does the work, so the two are not a fixed ratio. Because the models run on our own hardware rather than through a third-party API billed by the output second, the cost is our own measured compute, so we price it from what we actually observe instead of passing through a rate someone else sets and can change. As a guide, a 15 second clip at 720p currently takes 60 to 90 seconds of generation time, so a Starter month is around 40 to 60 of them.

† Self-Hosted, what it means: the platform is deployed inside the customer's own network rather than as shared SaaS, either on the customer's own hardware (on-prem) or in their private or government cloud tenancy, reachable from the intranet only, up to fully air-gapped installations with no internet connectivity. All models ship with the deployment and run locally; no third-party APIs are called, and no prompts, content, or telemetry ever leave the customer's environment. Updates arrive as offline packages, on your schedule. Packages are signed (coming soon). Quotas are unlimited because generation runs on customer-owned GPUs, so capacity is the deployment's hardware rather than shared infrastructure. Storage is sized to your hardware, up to 1 PB per deployment; larger by arrangement.

Processing tiers: your place in the queue, not render speed. To keep delivery fast for everyone, each account runs a limited number of generations at the same time, so a very large batch queues behind itself and takes longer to finish rather than jumping ahead of other customers. Turnaround estimates describe individual jobs as they start, not an entire batch.

Monthly pools, on your own clock: every allowance is a monthly pool that resets on your personal billing date, the day you signed up or upgraded, not the calendar month. There are no daily limits; skip a quiet week, or spend a month's minutes in one afternoon. When a trial month ends, your work stays viewable and downloadable; pick a plan to keep creating. An account with no paid plan closes after 12 months with no sign-in and its content is erased, and we email you before that happens.

What a minute costs: per minute of finished 1080p video that is 20 cents on Starter, 10 cents on Pro and 8 cents on Business. Generation runs on our own GPUs with open models and no per-call API fees, which is why the numbers are a fraction of credit-based tools; the full comparison is below.

Annual billing, all paid tiers: "Sign up for the whole year, get two months free." Starter $120, Pro $250, Business $500 per seat per year; exactly two months free on every tier. Monthly price is always the headline price.

Private beta: the beta program is first come, first served, closing at 1000 user sign-ups, with a public waitlist beyond that. During the beta you get privileged access in exchange for your patronage and feedback, and the option to sign up with the Pro plan for life at $125/year instead of $250, for as long as the subscription stays active. Cancel anytime; access runs through the end of the paid year.

Join the private beta

We email you when your slot opens. No spam, no drip sequences. Read how we handle your email in our Privacy Policy.

Questions about the online plans

Signing in without a password, what happens when you delete something, and whether anything you make is ever used to train a model.

See the FAQ →

How the market compares

We publish the entire market's prices next to ours, including every competitor's tiers, what a credit actually buys, and which vendors publish what a second of video costs.

See the full pricing comparison →