Rate limits

Request and concurrency limits and recommended polling intervals.

Base URL: https://api.ail.run

OpenAPI 3.1 definition · Full reference for language models

Contents

• Quickstart — Create an API key and generate a first Seedance video with curl, Python or Node.js.
• Authentication — Bearer API keys and their controls: model allowlists, spending limits, expiry, rate and concurrency limits and IP allowlists.
• Models — Seedance models, generation modes, resolutions, durations, aspect ratios and input limits.
• Pricing — How video tokens are counted (width × height × frames ÷ 1024), reserved at submission, settled and released.
• Uploading assets — Upload images, video and audio, or import them by URL, into the private asset library.
• Portraits — Virtual portraits and authorized real-person portraits for reference-to-video generation.
• Tasks — Create, query, list and cancel video generation tasks; task states and results.
• Idempotency — Retry task creation safely with the Idempotency-Key header.
• Result retention — How long generated videos are kept and how to get a fresh download link.
• Webhooks — Final-state webhooks signed with HMAC-SHA256 (X-AILink-Signature), delivered at least once with retries.
• Errors — Stable error codes, HTTP statuses and what to do about each.
• Rate limits — Request and concurrency limits and recommended polling intervals.
• Changelog — Changes to the API and the model catalog.
• Ark-compatible — Keep the official Volcengine Ark SDK and point base_url at this site.
• OpenAI-compatible — Create and fetch videos with the Videos API of the OpenAI SDK.
• Asset API (AK/SK) — Volcengine-compatible asset management API signed with an access key pair.