Versely

    Home/Pricing/Enterprise

    Versely Dedicated

    Unlimited AI video and UGC, on a GPU that is only yours

    For agencies and brands making hundreds or thousands of videos a month. We host the video model on hardware reserved for your team, so there are no per-video credits to run out of. What you can make is set by what the GPU can render, and we size that to your volume before you sign.

    What you get

    A GPU that renders only your work
    We host LTX 2.5, the open-weights video model behind Versely's AI hooks, on GPUs reserved for your team. Nothing of yours waits behind another customer's queue.
    No credits on the dedicated model
    Generate as many AI videos, UGC-style clips and hooks as the hardware can render. The limit is the GPU's throughput, not a balance that runs out mid-campaign.
    Sized to your volume, before you sign
    Tell us how many videos you make in a month. We work out how many GPUs that needs and the queue time to expect at that load, and you see both numbers before you commit.
    The whole Versely product around it
    Brand kits, the campaign planner, the video editor, captions, dubbing, scheduling and publishing, on the web and in the app, for every seat on your team.

    Who it fits

    A good fit

    • Agencies producing UGC-style ads for several clients every week
    • Ecommerce brands testing dozens of creatives per product
    • Apps and publishers generating short video at a steady daily volume
    • Teams that want a fixed monthly cost they can put in a budget

    Probably not yet

    • A few videos a week - a normal plan or a credit pack costs less
    • Work that must be made on a specific third-party model every time
    • Anyone who needs a contractual uptime guarantee today

    Making less than that? The regular plans and credit packs will cost you less.

    How it starts

    1. Step 1

      Send the form

      Monthly volume, what you make, and who it is for. Two minutes.

    2. Step 2

      Get a sized quote

      We reply by email with the GPU count for your volume, the expected queue time, and a monthly price.

    3. Step 3

      Run a paid pilot

      Your team makes real work on the dedicated GPU before anyone agrees to a longer term.

    Questions

    What does unlimited actually mean?
    No per-video credits on the dedicated model. You can queue as much as you like, and the GPU renders it in order. A GPU can only render so many seconds of video in an hour, so heavy days mean a longer queue, not a bill. If the queue is regularly longer than you want, the fix is another GPU, and we will tell you when that point is near.
    Which model runs on the dedicated GPU?
    LTX 2.5, an open-weights video model we already host ourselves for Versely's AI hooks. It is fast and strong at short vertical clips, product b-roll and hook-style openers. We will send you sample output for your own product before you pay for anything, so you can judge the quality yourself.
    Can we still use Veo, Seedance, Kling and the other models?
    Yes. Every model in the Versely catalog stays available on the same account, paid with credits as usual. Those providers charge per generation, so they cannot be part of a flat plan. Most teams use the dedicated model for volume and the premium models for hero pieces.
    How is it priced?
    Per dedicated GPU, per month, quoted after we know your volume. There is no public price because the right number of GPUs differs a lot between a team making 500 videos a month and one making 10,000.
    Is there an uptime guarantee?
    Not a contractual one yet. We run the same stack for our own AI hooks every day and we will tell you plainly what has and has not gone wrong with it. If a guaranteed SLA is a hard requirement for you, say so in the form and we will be honest about whether we can meet it.
    Who owns what we make, and is our footage used for training?
    The same terms as any Versely plan apply: you keep ownership of the content you submit, and we do not use your content to train models of our own. The dedicated GPU runs a fixed open-weights model; nothing you make changes it.
    Do you offer an API?
    Versely has an API and a CLI today. If you want to drive the dedicated GPU from your own system, tick API access in the form and describe the integration, and we will scope it with the quote.