Adlark

API reference

Partner API

REST over HTTPS. JSON in and out, multipart for uploads. Synchronous except video.

Authentication

Create a key under Account. Send it on every request:

Authorization: Bearer adl_...

Base URL: https://adlark.alexshipit.com. Files in responses (/api/files/...) need the same header. Files are kept 60 days.

Pricing and limits

Each call costs the studio credit price plus 10%, rounded up, from the same balance. Failed calls are refunded. Limits: 120 requests a minute per key, 20 generations a minute per account, and a daily credit cap per account (write to us to raise it).

Errors

Errors are JSON { "error": "..." }: 400 bad input, 401 bad key, 402 not enough credits, 413 file too large, 429 rate limited, 502/503 provider trouble (credits returned).

Quick start

curl -s https://adlark.alexshipit.com/api/v1/images \
  -H "Authorization: Bearer $ADLARK_API_KEY" \
  -H "content-type: application/json" \
  -d '{"prompt":"Studio photo of a ceramic mug, warm light","model":"klein","aspect":"1:1"}'

Endpoints

GET /api/v1/credits

Credit balance and rank.

{ "credits": 2950, "plan_credits": 2950, "pack_credits": 0, "rank": "bronze" }

POST /api/v1/images

Generate 1-4 images. JSON with base64 reference_images, or multipart with `refs` files.

{ "prompt": "Amber serum bottle on wet stone, morning light", "model": "klein", "size": "1K", "aspect": "4:5", "count": 2, "reference_images": ["<base64>"] }
{ "job": "Jb3...", "credits": 18, "images": [{ "id": "...", "url": "/api/files/...", "width": 816, "height": 1024 }], "failed": 0 }

POST /api/v1/copy

Ad copy set for one platform.

{ "product": "Glowline serum", "details": "15% vitamin C, fragrance free", "audience": "women 25-40", "offer": "20% off", "platform": "TikTok", "tone": "Friendly", "language": "English" }
{ "job": "...", "credits": 6, "copy": { "angles": [...], "hooks": [...], "headlines": [...], "primary_texts": [...], "descriptions": [...], "ctas": [...] } }

POST /api/v1/ugc-script

Scene-by-scene UGC script. Same brief fields plus length (15, 30, 45, 60), creator, angle.

{ "product": "Glowline serum", "length": 30, "creator": "busy mom in her 30s" }
{ "job": "...", "credits": 6, "script": { "title": "...", "scenes": [{ "n": 1, "time": "0-3s", "shot": "...", "voiceover": "...", "on_screen_text": "...", "image_prompt": "..." }], "hook_variations": [...] } }

POST /api/v1/scenes

Split a script into 3-8 scenes with image prompts.

{ "script": "...", "style": "bright handheld look" }
{ "job": "...", "credits": 6, "scenes": [...] }

POST /api/v1/analyze

Multipart: `file` (image or video up to 3 min / 80 MB), `caption`, `platform`.

curl -F file=@ad.mp4 -F platform=TikTok ...
{ "job": "...", "credits": 55, "analysis": { "grade": "B", "score": 74, "hook": {...}, "cta": {...}, "audience": {...}, "ugc_script": {...}, "hook_variations": [...] }, "transcript": { "text": "..." }, "frames": [{ "url": "...", "at": 0.5 }] }

POST /api/v1/captions

Multipart: `file` (video), `style` (bold-pop | clean | boxed | neon | karaoke), `uppercase` (1/0).

curl -F file=@edit.mp4 -F style=bold-pop ...
{ "job": "...", "credits": 28, "video": { "url": "/api/files/..." }, "srt": { "url": "/api/files/..." } }

POST /api/v1/voiceover

Text to MP3. Voices: thalia, andromeda, helena, apollo, arcas, orion.

{ "text": "Meet the serum that changed my mornings.", "voice": "thalia" }
{ "job": "...", "credits": 2, "audio": { "url": "/api/files/..." }, "characters": 41 }

GET /api/v1/jobs/{id}

A job's status and output (video jobs are polled here).

{ "id": "...", "tool": "image", "status": "done", "credits": 18, "output": {...} }

POST /api/v1/videos

Coming soon: returns 503 until video launches. Then: prompt, model (veo31fast | veo31), seconds (4|6|8), aspect, resolution, first_frame (multipart).

{ "job": "...", "credits": 282, "status": "running" }

Agents: a plain-text index is at /llms.txt. MCP: see /mcp.