GET STARTED

Lithovas API

Access chat, image, video, audio, and embedding models through one authenticated API.

Lithovas gives developers one account, one API key, and one credit wallet for multimodal generation. Use the OpenAI-compatible endpoint for chat, or submit asynchronous media tasks and poll until the result is ready.

Run your first request

Create a key and complete a request from submission to result.

→

Get an API key

Create and manage server-side credentials in Lithovas Console.

↗

Make a request

Every protected endpoint uses a Bearer API key. Chat returns an OpenAI-compatible response; media generation returns a task identifier that you can poll for status and output.

curl --request POST \
  --url https://api.seedance.ch/v1/videos \
  --header "Authorization: Bearer ${LITHOVAS_API_KEY}" \
  --header "Content-Type: application/json" \
  --data '{
    "model": "seedance-2.5-text-to-video",
    "prompt": "A cinematic product reveal",
    "seconds": "5",
    "size": "1280x720"
  }'

curl --request POST \
  --url https://api.seedance.ch/v1/chat/completions \
  --header "Authorization: Bearer ${LITHOVAS_API_KEY}" \
  --header "Content-Type: application/json" \
  --data '{
    "model": "chat/qwen3-7-plus",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": false
  }'

The media endpoint immediately returns task information. Save the task ID and query it until the task reaches a terminal state.

{
  "taskId": "task_1723996800000_abc123",
  "status": "pending"
}

Authentication

Send your API key in the Authorization header. Keep keys on your server and never expose them in browser code or public repositories.

Authorization: Bearer $LITHOVAS_API_KEY

Choose your path

{ }Build with chatUse the OpenAI SDK, streaming, messages, and vision inputs.→ ▷Generate mediaCreate asynchronous image, video, and audio tasks.→ ↑Upload reference mediaUpload a file first, then pass the returned HTTPS URL to a model.→

Billing behavior Failed media generation tasks are not billed. Download completed output that you need to retain; provider result URLs should not be treated as permanent storage.

API REFERENCE

Chat completions

Call supported chat and vision models with the OpenAI-compatible Chat Completions API. Non-streaming responses use the standard choices[0].message.content shape and include Lithovas credit usage.

POSThttps://api.seedance.ch/v1/chat/completions

Media generation

Submit video tasks with top-level model, prompt, seconds and model-specific parameters. Image generation uses POST https://api.seedance.ch/v1/images/generations.

POSThttps://api.seedance.ch/v1/videos

Task status

Poll the authenticated task endpoint with the ID returned when the media task was created.

GEThttps://api.seedance.ch/v1/videos/{taskId}

Uploads

The signed-in web playground uploads raw image, video or audio bytes with the matching Content-Type and X-File-Name header. This website-session endpoint returns a temporary HTTPS reference URL valid for 24 hours; it is not an API-key upload endpoint.

POSThttps://www.seedance.ch/api/open/media-upload

Embeddings

Use the OpenAI-compatible embeddings request format with a model enabled for your API key. Check the current model catalog before submitting.

POSThttps://api.seedance.ch/v1/embeddings

Model catalog

Discover the platform's currently available model slugs, providers, task types, and input capabilities before constructing requests.

GEThttps://api.seedance.ch/v1/models