One OpenAI-compatible API, one wallet, one dashboard. New models are added continuously —
list them anytime with GET /v1/models.
Led by the new Seedance 2.5 flagship — multimodal, up to 30s — with Seedance 2.0 (to 4K) and speed- and budget-tuned tiers. Pick quality, speed and cost per job.
The new flagship. Multimodal text / image / video / audio → video with longer narratives, stronger reference control, precise editing & extension, and more realistic motion, lighting and performance.
The previous frontier model — text- and image-to-video up to 4K with native audio.
The same core model tuned for speed — ideal for iteration, previews, and high-volume pipelines.
The most affordable way to prototype cinematic AI video at scale.
Video is billed per output second, per model & resolution. Call GET /v1/models for live per-model pricing.
Long-context conversational models over the same OpenAI-compatible endpoint — swap the base URL and keep your SDK.
256K-context conversational models over the same OpenAI-compatible endpoint. Token-by-token streaming — point your SDK’s base_url at us and change nothing else.
Text-to-image and image editing, served the same drop-in way.
Text-to-image and image editing at 1K or 2K via /v1/images/generations — flat rate at either size.
Free points on signup — no card required to start. Read the docs or grab a key and ship today.
