Seedance 2 Fast - Kyma API

Overview

Developed by ByteDance, this model generates video and native audio from text or image prompts. It operates as the fast tier of the Seedance 2 family, delivering quicker generation times at a lower cost than the Pro variant. On Kyma, the endpoint is OpenAI-compatible and routes through automatic failover if a serving path degrades. You can enable prompt caching to bill repeated prefixes at this model’s cached input rate, and every response returns exact billing in usage.cost alongside the X-Kyma-Model header. New accounts receive a $0.50 credit to start testing immediately. It does not support vision input analysis, structured outputs, or reasoning chains. The model is billed per second of generated output and is optimized for shorter clips rather than long-form video.

Specs

Field	Value
Model ID	`seedance-2-fast`
Best for	Social shorts, UI motion, rapid iteration, product demos
Context window	0
Input modalities	Text, Image
Output modalities	Video, Audio
Tool calling	Yes
Structured outputs	Yes
Prompt caching	Yes
Speed	Fast
Cost band	Balanced
Release stage	Stable

Pricing

	Per 1M tokens
Input	$0.20
Output	$0.60

Use this when

Social Media Shorts — Generate quick, engaging video clips with bundled audio for platforms like TikTok or Reels.
UI Motion Prototyping — Create animated mockups and interface transitions from static design assets.
Rapid Product Demos — Iterate on short promotional videos or feature walkthroughs with fast turnaround times.

Not ideal for

Do not use this model for long-form cinematic video, complex multi-step reasoning tasks, or workflows requiring structured JSON outputs.

Pick something else when

You need higher visual fidelity and longer clip duration: use seedance-2-pro.
You require advanced cinematic realism and extended generation times: use kling-3-pro or veo-3.
You need fast generation with higher resolution output: use hailuo-02-1080p.

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-...")

response = client.chat.completions.create(
    model="seedance-2-fast",
    messages=[{"role": "user", "content": "..."}],
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=generate-video&tools=true&quality_tier=strong"

FAQ

Does this model support image-to-video generation? Yes, it accepts both text and image inputs to produce video and native audio outputs. How does prompt caching work for video generation? Repeated prompt prefixes are billed at this model’s cached input rate, reducing costs for iterative workflows. What happens if a generation request fails? Kyma automatically reroutes degraded requests through alternate serving paths, so a single bad path does not surface as a failed generation.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ