Skip to main content
View on GitHub → AI-powered media generation for autonomous agents. One AISA_API_KEY unlocks Gemini-compatible GenerateContent and Qwen Wan 2.6 video synthesis.

​
Install

First, install the AIsa CLI if you have not already: 
npm install -g @aisa-one/cli
Then install the skill: 
aisa skills install media-gen

​
What can agents do with it?

Marketing creatives

“Generate a cinematic hero image for a product launch deck.”

Social content

“Create a 5-second video loop for the next post.”

Storyboarding

“Produce 6 key frames illustrating the happy path of a user journey.”

Reference-to-video

“Animate this static mock into a cinematic slow push-in.”

Photorealism

“8k ultra-detailed cyberpunk skyline with neon rain.”

Agent reports

“Generate visual illustrations for a research report agent.”

​
Core capabilities

  • Gemini-compatible generation — use POST /v1beta/models/{model}:generateContent with a verified model such as gemini-3.5-flash for native Gemini-format content generation.
  • Video generation — wan2.6-t2v (text-to-video) and image-to-video via an async task system. POST creates a task, GET polls status.
  • Asynchronous workflow — long-running video jobs are handled with X-DashScope-Async: enable + status polling.

​
Quick start

export AISA_API_KEY="your-key"

​
Gemini-compatible generation

curl -X POST "https://api.aisa.one/v1beta/models/gemini-3.5-flash:generateContent" \
  -H "Authorization: Bearer $AISA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {"role": "user", "parts": [{"text": "Return exactly: ok"}]}
    ],
    "generationConfig": {"maxOutputTokens": 16}
  }'
The response follows the Gemini GenerateContentResponse shape, with generated text in candidates[0].content.parts[0].text.

​
Video generation (async)

# 1. Create the task
curl -X POST "https://api.aisa.one/apis/v1/services/aigc/video-generation/video-synthesis" \
  -H "Authorization: Bearer $AISA_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-DashScope-Async: enable" \
  -d '{
    "model": "wan2.6-t2v",
    "input": {
      "prompt": "cinematic close-up, slow push-in, shallow depth of field",
      "img_url": "https://example.com/reference.jpg"
    }
  }'

# Response includes a task_id. Poll for completion:
curl "https://api.aisa.one/apis/v1/services/aigc/tasks?task_id=YOUR_TASK_ID" \
  -H "Authorization: Bearer $AISA_API_KEY"
When the task status is SUCCEEDED, the response includes a video_url you can download.

​
Python client

# Video — create + poll + download in one command
python3 scripts/media_gen_client.py video-wait \
  --prompt "cinematic close-up, slow push-in" \
  --download --out out.mp4

# Or create/poll separately
python3 scripts/media_gen_client.py video-create --prompt "cinematic sunset"
python3 scripts/media_gen_client.py video-status --task-id YOUR_TASK_ID

​
Endpoint reference

EndpointMethodPurpose
/v1beta/models/{model}:generateContentPOSTGemini-compatible generation
/apis/v1/services/aigc/video-generation/video-synthesisPOSTCreate video task
/apis/v1/services/aigc/tasksGETPoll video task status

​
Get started

  1. Sign up at aisa.one (new accounts start with $2 free credit).
  2. Generate an API key from the console.
  3. export AISA_API_KEY="your-key" and install the skill: 
    npm install -g @aisa-one/cli
aisa skills install media-gen

Video API reference

Create-task and poll-status endpoints with live playgrounds.

Gemini generateContent

Gemini-compatible generation endpoint reference.

Async Operations

How task creation and polling work end-to-end.