Skip to Content

API Stock

API Stock is an aggregator for AI generation models. Instead of holding an account, a key and a billing relationship with every vendor whose model you want to use, you integrate once and reach all of them through a single endpoint, a single API key and a single prepaid balance.

  • 27 media models — video, image and music, from Sora 2, Veo 3.1, Nano Banana, Seedance, Seedream, Kling, Pixverse, Wan 2.7, Midjourney and Suno.
  • 3 chat models over the OpenAI and Anthropic wire protocols, so the official SDKs work against us with nothing but a changed base URL.
  • Automatic failover — every model is served by several upstream providers. If one fails, the same generation is retried on a reserve provider before you ever see an error.
  • Results on our storage — finished media is downloaded from the provider and re-hosted by us, so a result URL does not disappear when the upstream vendor expires it.
  • Prepaid, per-request billing in whole cents, refunded automatically on every failure path.

Your first request

curl -X POST https://api.api-stock.com/api/v1/generation/create \
  -H "Authorization: Bearer $API_STOCK_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "veo3.1-fast",
    "input": { "prompt": "A waterfall cascading into a rainbow", "resolution": "1080p" }
  }'

The call returns immediately with a task id — generation is asynchronous. Poll GET /task/status/{taskId} or supply a webhook URL to be notified when it finishes.

Where to next

  • Quick start — get a key and run a generation end to end.
  • Authentication — API keys, session cookies and the two headers we accept.
  • Generation lifecycle — what happens between not_started and finished, and how retries and refunds work.
  • Models — the catalogue, what each model takes, and the parameters they share.
  • LLM & chat — the OpenAI- and Anthropic-compatible chat endpoints and how tokens are billed.
  • API Reference — the interactive reference, generated from the server’s OpenAPI document.