Superseedance Guides MCP Server

SuperSeedance Guides

Practical Seedance video guides, complete prompts and original examples from SuperSeedance.

简体中文

SuperSeedance is an independent AI video workspace for creating from text or images with Seedance 2.5. It brings prompts, private reference images, credit estimates and saved videos into one place, with an interface in 20 languages.

Start with your shot

Watch the original examples

The prompt guide pairs eight original Seedance 2.5 clips with their complete prompts and generation settings: rainy-night parkour, a cloud-city flight, anime skating, an ink-wash whale, solo dance, clay animation, a forest adventure and liquid metal.

The submitted settings for these eight examples were text-to-video, six seconds, 720p, 16:9, no audio and no draft mode. Preview edits do not change those source-generation settings. A prompt is a reproducible starting point; it does not guarantee an identical output.

Using SuperSeedance

Choose text or image input, describe one scene and its movement, select supported settings, review the studio-credit quote, then submit and track the task in My Creations. Generation requires a verified account, sufficient credits and an available service.

Credit purchases are not yet available as of October 24, 2026. Existing guides and public examples remain accessible. Check the pricing page for the current purchase status.

About

This documentation is maintained for SuperSeedance. The platform is independently operated by 幸会新知科技（深圳）有限公司 and is not affiliated with ByteDance. This repository contains guides; the hosted application source code is not included.

About · Contact · Terms · Privacy

Sample videos are presented for demonstration; viewing this repository does not grant a commercial redistribution license for them. Use your own or authorized input material and review the applicable terms before publishing generated work.

The product-ad wording preview and expanded optional measurement are live. The account-initialization input issue was fixed and passed eight additional live checks. The 856-test suite, typecheck, build and three isolated CTA renders passed without new generation jobs or credit entries. Local and production English/Chinese browser checks passed at 320/360/390/430/1280px without horizontal overflow; the production check retained an existing signed-in session, used measurement exclusion and submitted no upload, generation or ad assembly. Acceptance aggregate counts showed no new assets, jobs, ads or credit spending. These were viewport checks, not physical-device tests; the three renders were local only. See the product-ad validation scope.

Updated October 27, 2026. The image-diagnostics tool is live, with both first-frame and Reference preparation flows checked online without uploading an image or submitting a generation. Local checks and signed-in mobile-width browser checks passed, with no new saved assets, generation tasks or credit spending during acceptance. The checks used resized browser viewports, not physical iOS/Android devices or a real-user study. The live website is the source for current availability and prices.

The English dialogue preparation tool is live and passed local and online free-flow and resized-phone-viewport checks. It prepares A/B prompts and a manual review plan without generating audio. Both creator handoffs were checked; acceptance aggregates showed no new jobs or credit spending. No audio sample, listening result or repair rate is claimed. See the validation scope.