PERSEWORKS
// faq

Frequently asked questions

Everything you need to know before you build. Still stuck? Read the docs or email support@perseworks.hk.

Is this really Seedance? Which version? +

Yes — every video is generated by ByteDance’s Seedance 2.0 model, unmodified. We’re a gateway that adds an OpenAI-compatible API, USD billing, and content moderation on top of the raw model.

Do I need a Chinese account, phone, or bank? +

No. Sign up with an email, pay with a card in USD, and get an API key in about 60 seconds. Removing that friction is the whole point.

How is this different from BytePlus, Runway, or Sora? +

BytePlus is ByteDance’s official platform — powerful, but built around cloud-account onboarding and its own SDK/signature auth. We give you the same frontier Seedance model with a 60-second, OpenAI-compatible, pay-as-you-go setup and no enterprise sales process. Versus Western tools like Runway or Higgsfield, you get frontier Chinese video, usually cheaper, with a drop-in API.

Can I use the output commercially? +

Generally yes, subject to the upstream model provider’s usage terms and your own responsibility to secure any rights or consents needed for real people, trademarks, or copyrighted material referenced in your prompts. You are responsible for how you use generated content and for any AI-content labeling required in your market.

Is there a watermark on the video? +

No — output is delivered without a platform watermark.

Text-to-video and image-to-video? +

Both. Send a prompt for text-to-video, or include a reference image URL to drive image-to-video.

How long does a video take to generate? +

Typically tens of seconds to a couple of minutes depending on length and resolution. Generation is asynchronous: your submit call returns a task id immediately, then you poll for the result.

What resolutions and durations are supported? +

480p, 720p, and 1080p, up to 15 seconds per clip. Longer or higher-throughput needs can be discussed — contact us.

How does billing work? +

Prepaid credits (1 credit = $0.01), no subscription. Video is billed per output second by resolution and charged for the actual duration generated; chat is billed per token. New accounts get free trial credits.

Are there rate limits? +

Yes — sensible per-key request limits protect the platform from abuse. If you need higher throughput for a production workload, contact us to raise them.

What is your refund policy? +

Credits are non-refundable once purchased, except as required by law or for duplicate / erroneous charges. See the Refund Policy for details.

Where is my prompt processed? Is my data private? +

To generate output, your prompt (and any reference media) is sent to the model provider (ByteDance / Volcano Ark) and processed in mainland China, then the result is returned to you. We don’t persistently store request inputs, except content flagged by moderation for review. See the Privacy Policy.

Is it really OpenAI-compatible? +

Chat is a drop-in replacement — point the OpenAI SDK’s base_url at us and change nothing else. Video uses a simple submit-and-poll REST endpoint documented in the Docs.

Do you offer a free trial? +

Yes — free trial credits on signup, no card required to start.