> ## Documentation Index > Fetch the complete documentation index at: https://www.tryleap.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Leap's API is at https://api.tryleap.ai. Send the key in the x-api-key header or as Authorization: Bearer; keys start with leap_. > Take model IDs and input schemas from GET /v1/models/{creator}/{name}, or from https://api.tryleap.ai/v1/public/models without a key. Never guess an input field: unknown fields are a 400. > Model inputs take uploaded file IDs (file_...), never URLs. > To give a coding agent the whole API, install the Leap skill: curl -fsSL https://www.tryleap.ai/install.sh | sh > The full OpenAPI 3.1 spec is at https://www.tryleap.ai/docs/openapi.json. # Quickstart > Create an API key, make an image in one request and save it, then start a video and poll it until it's done. No key yet? Browse models and schemas at [https://api.tryleap.ai/v1/public/models](https://api.tryleap.ai/v1/public/models). It needs no key, and it lists every model ID with its price and input schema. Open [app.tryleap.ai/go/api](https://app.tryleap.ai/go/api), sign in, and create a key. New workspaces start with \$2 of free credit, and the examples on this page cost about \$0.69 in all, so you can try them before you add a card. Limits against abuse can withhold the credit, for example after several sign-ups from one network on the same day. ```bash theme={"theme":"css-variables"} export LEAP_API_KEY="leap_..." ``` Keep the key on your server. It spends your workspace's balance. `prefer: wait=60` holds the request open until the image is ready, for up to 60 seconds, so one request gets you the result. To run the Node.js version, save it as `quickstart.mjs` and run `node quickstart.mjs` (Node 18 or later). For the Python version, run `pip install requests` first. The curl version needs [jq](https://jqlang.org). ```bash curl theme={"theme":"css-variables"} RESPONSE=$(curl -s https://api.tryleap.ai/v1/generations \ -H "x-api-key: $LEAP_API_KEY" \ -H "content-type: application/json" \ -H "prefer: wait=60" \ -d '{ "model": "black-forest-labs/flux-2-pro", "input": { "prompt": "A silk scarf in the wind on a ferry deck", "aspect_ratio": "3:4" } }') echo "$RESPONSE" | jq . ``` ```js Node.js theme={"theme":"css-variables"} import { writeFile } from "node:fs/promises"; const response = await fetch("https://api.tryleap.ai/v1/generations", { method: "POST", headers: { "x-api-key": process.env.LEAP_API_KEY ?? "", "content-type": "application/json", prefer: "wait=60", }, body: JSON.stringify({ model: "black-forest-labs/flux-2-pro", input: { prompt: "A silk scarf in the wind on a ferry deck", aspect_ratio: "3:4" }, }), }); const generation = await response.json(); if (!response.ok) throw new Error(generation.error.message); if (generation.status === "succeeded") { const { url } = generation.output[0]; console.log(url); // The link is signed, so downloading it needs no API key. const image = await fetch(url); if (!image.ok) throw new Error(`Download failed: ${image.status}`); await writeFile("scarf.jpg", Buffer.from(await image.arrayBuffer())); } else if (generation.status === "failed") { console.error(generation.error.message); process.exitCode = 1; } else { // Still queued or running when the wait ended: poll it until it's done. console.log(`${generation.id} is ${generation.status}. Poll GET /v1/generations/${generation.id}.`); } ``` ```python Python theme={"theme":"css-variables"} import os import requests response = requests.post( "https://api.tryleap.ai/v1/generations", headers={"x-api-key": os.environ["LEAP_API_KEY"], "prefer": "wait=60"}, json={ "model": "black-forest-labs/flux-2-pro", "input": {"prompt": "A silk scarf in the wind on a ferry deck", "aspect_ratio": "3:4"}, }, timeout=90, ) generation = response.json() if not response.ok: raise RuntimeError(generation["error"]["message"]) if generation["status"] == "succeeded": url = generation["output"][0]["url"] print(url) # The link is signed, so downloading it needs no API key. image = requests.get(url, timeout=60) image.raise_for_status() with open("scarf.jpg", "wb") as f: f.write(image.content) elif generation["status"] == "failed": raise SystemExit(generation["error"]["message"]) else: # Still queued or running when the wait ended: poll it until it's done. print(f"{generation['id']} is {generation['status']}. Poll GET /v1/generations/{generation['id']}.") ``` The answer is a generation. Its `output` holds signed links to the files: ```json theme={"theme":"css-variables"} { "id": "gen_4cB8rTq2LmX9vK1nW7pZs3Ad", "object": "generation", "model": "black-forest-labs/flux-2-pro", "modality": "image", "status": "succeeded", "input": { "prompt": "A silk scarf in the wind on a ferry deck", "aspect_ratio": "3:4", "n": 1 }, "output": [ { "type": "image", "url": "https://api.tryleap.ai/files/eyJrIjoibWVkaWEv...", "expires_at": "2026-10-05T18:22:47.120Z", "content_type": "image/jpeg", "width": 896, "height": 1152 } ], "error": null, "usage": { "cost_usd": "0.033" }, "created_at": "2026-10-04T18:22:41.512Z", "completed_at": "2026-10-04T18:22:47.120Z" } ``` Anyone with the `url` can download the file until `expires_at`, without an API key. Links expire after 24 hours, so download what you keep. The Node.js and Python scripts above save the image as `scarf.jpg`. With curl: ```bash theme={"theme":"css-variables"} curl -L -o scarf.jpg "$(echo "$RESPONSE" | jq -r '.output[0].url')" ``` `-L` follows the redirect to storage that large files, such as videos, get. Reading the generation again with `GET /v1/generations/{id}` signs fresh links. The full object is in [Generations](/docs/generations#the-generation-object). Video takes a minute or more, longer than one request can wait. This script starts a clip without waiting, then long-polls it: each read waits up to 60 seconds for the clip to finish. It gives up after 15 minutes, and saves the clip as `heron.mp4`. Save it as `video.sh` and run `bash video.sh`. It needs jq 1.6 or later. ```bash theme={"theme":"css-variables"} #!/usr/bin/env bash set -euo pipefail API=https://api.tryleap.ai RESPONSE=$(curl -sS "$API/v1/generations" \ -H "x-api-key: $LEAP_API_KEY" \ -H "content-type: application/json" \ -d '{"model": "google/veo-3.1-fast", "input": {"prompt": "A heron lands on a still lake at dawn", "duration": 4}}') # An error answer has no id: print its message and stop. GENERATION_ID=$(jq -r '.id // empty' <<< "$RESPONSE") if [ -z "$GENERATION_ID" ]; then jq -r '.error.message // .' <<< "$RESPONSE" >&2 exit 1 fi echo "Started $GENERATION_ID" DEADLINE=$(( $(date +%s) + 15 * 60 )) while true; do # Each read waits up to 60 seconds for the run to end. A dropped request or an # error answer leaves STATUS empty, and the loop reads again. GENERATION=$(curl -sS --max-time 75 "$API/v1/generations/$GENERATION_ID" \ -H "x-api-key: $LEAP_API_KEY" -H "prefer: wait=60" || true) STATUS=$(jq -r '.status // empty' <<< "$GENERATION" 2> /dev/null || true) case "$STATUS" in succeeded) break ;; failed | canceled) echo "$STATUS: $(jq -r '.error.message // "no message"' <<< "$GENERATION")" >&2 exit 1 ;; esac if [ "$(date +%s)" -ge "$DEADLINE" ]; then echo "$GENERATION_ID is still running. Read it again later." >&2 exit 1 fi sleep 2 done curl -sSL -o heron.mp4 "$(jq -r '.output[0].url' <<< "$GENERATION")" echo "Saved heron.mp4" ``` This 4-second clip with sound costs \$0.66. [Generations](/docs/generations#long-poll) shows the same long-poll in TypeScript and Python. ## Next Browse the catalog and read what each model takes. Upload a file, then animate or edit it. Upload a product photo, animate it with Veo and save the clip. Start runs from a route handler and show the result on a page. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.