AIVeed Developers Get your API key

Use AIVeed with your AI assistant

AIVeed is built to be driven by LLMs: the whole API fits in one plain-text file — https://developer.aiveed.io/llms-full.txt — pricing is queryable at GET /v1/credits/pricing (public), and every response is available in the token-frugal TOON format. Paste a pack below into Claude, ChatGPT, or any assistant that can fetch URLs and run code.

Set AIVEED_API_KEY in the assistant's environment first — never paste the key into the chat itself. Get one via Get your API key above.

Generate a video

prompt
Fetch https://developer.aiveed.io/llms-full.txt and read it. You are now my AIVeed API assistant.
My API key is in the environment variable AIVEED_API_KEY.
Generate a 10s 9:16 Kling 3.0 (model "kling-3.0", mode "pro") video of <YOUR PROMPT HERE>,
wait for completion by polling GET /v1/jobs/{id} every 10 seconds,
then give me the hosted URL and how many credits it cost.

Per-model variants

Veo3 — generate then extend

prompt
Fetch https://developer.aiveed.io/llms-full.txt and read it. You are now my AIVeed API assistant.
My API key is in env AIVEED_API_KEY.
Generate a 9:16 video with model "veo3-fast" of <YOUR PROMPT HERE> (duration is fixed ~8s — do not send a duration).
When it completes, extend it once via POST /v1/extend/video with a continuation prompt you write yourself.
Report both hosted URLs and the total credits charged.

Seedance — cheapest video, price-checked first

prompt
Fetch https://developer.aiveed.io/llms-full.txt and read it. You are now my AIVeed API assistant.
My API key is in env AIVEED_API_KEY.
Generate the cheapest possible video: model "seedance-v1-pro", 480p, 5s, 9:16, of <YOUR PROMPT HERE>.
Check GET /v1/credits/pricing first and tell me the cost BEFORE submitting, then submit, poll, and give me the hosted URL.

Images

prompt
Fetch https://developer.aiveed.io/llms-full.txt and read it. You are now my AIVeed API assistant.
My API key is in env AIVEED_API_KEY.
Generate a product image with model "nano-banana-pro" of <YOUR PRODUCT DESCRIPTION>,
then create an upscaled variant by passing its result_url in image_urls to model "gpt-image-2" at resolution "2K".
Give me both hosted URLs and the credits charged for each.

Showcase: image → animate it

Exercises both modalities and the first-frame capability in one flow.

prompt
Fetch https://developer.aiveed.io/llms-full.txt and read it. You are now my AIVeed API assistant.
My API key is in env AIVEED_API_KEY.
Two-step demo, exercising both modalities:
1) Generate an image with model "nano-banana" (the cheapest image model) of <YOUR PROMPT HERE>. Poll until completed.
2) Animate it: pass the image's result_url as first_frame_url to model "seedance-2-fast" (480p, 6s) with a motion prompt you derive from the image prompt.
Poll to completion and give me the final video URL plus the total credits charged for both steps.

Build me an integration

prompt
Fetch https://developer.aiveed.io/llms-full.txt and read it — it describes the AIVeed API (AI video/image/audio generation, credit-priced, job-based).
Build me a minimal integration in <LANGUAGE/FRAMEWORK> that:
- reads AIVEED_API_KEY from the environment,
- exposes one function generateVideo(prompt, options) that submits POST /v1/generate/video, honors 429 Retry-After, and sends an Idempotency-Key,
- waits for completion via polling (or a webhook receiver that verifies the X-AIVeed-Signature HMAC, if I asked for webhooks),
- surfaces the error `code` values from the docs to the caller,
- never logs the API key.
Then show me a usage example with model "seedance-v1-pro" at 480p/5s.

Tips for agent builders