LivePair Examples

Working recipes for the LivePair AI API — image and video generation, chat completions, and x402 agent payments.

livepair.py is a dependency-free client (standard library only) that handles the full async lifecycle: submit → poll → media URL, with typed errors, timeouts, and one network retry. Drop it into any project.

from livepair import LivePair

client = LivePair()                       # reads LP_KEY from env
url = client.generate_image("a cat astronaut, cinematic")
clip = client.generate_video("slow wave crash, aerial")

Get a key at https://livepairai.com/settings — lp_…, prepaid credits, per-result billing. Wallet-carrying agents can skip the key entirely and pay per call with x402 (USDC on Base) — see x402-payment.md.

Examples

File	What it does
examples/generate_image.py	Generate an image, download it before the 48h TTL
examples/generate_video.py	Generate a clip with a longer poll timeout
examples/list_models.py	Print the live catalog — free, no key needed
mobile-proxy.mjs	Backend proxy for mobile apps — key stays server-side

export LP_KEY=lp_...
python examples/generate_image.py "a cat astronaut, cinematic"
python examples/generate_video.py "aerial drone shot of a coastline"
python examples/list_models.py image

The contract in 10 lines

curl -X POST https://livepairai.com/v1/agent/generate \
  -H "x-api-key: $LP_KEY" -H "Content-Type: application/json" \
  -d '{"model":"qwen-image-3","prompt":"a cat astronaut"}'
# → {"jobId": "..."}
curl https://livepairai.com/v1/agent/jobs/<jobId> -H "x-api-key: $LP_KEY"
# → {"status":"done","url":"https://…"}  — media auto-deletes after 48h

Errors you should handle: 401/403 (bad key), 402 (insufficient credit or an x402 quote), 5xx (retryable). livepair.py maps these to typed exceptions.

OpenAI-compatible surfaces

If your app already speaks OpenAI, skip the job protocol and point it at https://livepairai.com/v1/agent — three spec-shaped endpoints:

Endpoint	Spec	Notes
POST /v1/agent/chat/completions	OpenAI chat	model ids from GET /v1/agent/models
POST /v1/agent/images/generations	OpenAI Images	{prompt, model, size, response_format} → {created, data:[{url}]}
POST /v1/agent/videos	OpenAI Videos (Sora)	returns a job; GET /v1/agent/videos/{id} polls, …/content fetches the mp4

GET /v1/agent/models also serves the OpenAI {object:"list"} shape so client model pickers populate automatically. Same lp_ key (Bearer or x-api-key), same prepaid-credit billing.

Related

• livepair-cli — npm i -g livepair
• livepair-bots — Telegram/Discord/X templates
• n8n-nodes-livepair — n8n community node
• ComfyUI-LivePair — ComfyUI nodes
• livepair-skills — agent SKILL.md
• OpenAPI: https://livepairai.com/v1/agent/openapi.json · MCP: https://livepairai.com/mcp

MIT