# genkiln

> One API key for Veo, Seedance, Kling, Nano Banana and GPT Image. Pay as you go, failed tasks refunded.

genkiln is a REST API for AI image and video generation. One API key gives access to Veo 3.1 Lite, Veo 3.1 Fast, Seedance 2.0 Fast, Kling 3.0, Nano Banana 2, Nano Banana Pro and GPT Image 2. You create a task with `POST https://genkiln.com/v1/tasks`, it runs asynchronously, and you read the result by polling `GET /v1/tasks/{id}` or receive it on a signed webhook. Every model uses the same request shape.

- Base URL: `https://genkiln.com/v1`. Authenticate with `Authorization: Bearer sk_...` (create a key in the console after signing up).
- Billing: prepaid credits, 1 credit = US$0.001. No subscription or monthly fee, and credits never expire. The exact price of a task is fixed by our catalog and returned as `cost_credits` when you create it.
- A task that fails is refunded in full, automatically.
- Result files are kept for 30 days after the task completes, then deleted.
- Support: support@genkiln.com

## Docs

- [Quickstart](https://genkiln.com/docs/quickstart): Create an API key, submit your first image or video generation task, and fetch the result in a few minutes.
- [Authentication](https://genkiln.com/docs/authentication): Authenticate every genkiln API request with a secret API key sent as a Bearer token.
- [Create task](https://genkiln.com/docs/create-task): POST /v1/tasks: submit an image or video generation task, with request fields, validation rules, pricing and idempotency.
- [Get task](https://genkiln.com/docs/get-task): GET /v1/tasks/{id}: read a task, its status, output URLs and error. Reference for the task object and its statuses.
- [List tasks](https://genkiln.com/docs/list-tasks): GET /v1/tasks: page through your tasks newest first with cursor pagination, filtered by model or status.
- [Models](https://genkiln.com/docs/models): Every image and video model on genkiln: IDs, prices in credits and USD, and the input parameters each model accepts.
- [Webhooks](https://genkiln.com/docs/webhooks): Receive a signed POST when a task finishes: event types, headers, payload, retries and signature verification in Node.js and Python.
- [Errors](https://genkiln.com/docs/errors): The genkiln error format, every HTTP error code, task error codes, and which errors are safe to retry.
- [Rate limits](https://genkiln.com/docs/rate-limits): Request rate limits per API key, the concurrent task limit per account, request size limits and task time limits.
- [Changelog](https://genkiln.com/docs/changelog): Changes to the genkiln API, newest first.

## Models

- [Veo 3.1 Lite](https://genkiln.com/models/veo-3.1-lite): model ID `veo-3.1-lite`, by Google, text-to-video and image-to-video, from 195 credits ($0.195) per video. Google's most affordable Veo 3.1 tier: 4–8 second clips with native audio from a prompt or a first/last frame.
- [Veo 3.1 Fast](https://genkiln.com/models/veo-3.1-fast): model ID `veo-3.1-fast`, by Google, text-to-video and image-to-video, from 390 credits ($0.39) per video. Veo 3.1 Fast: cinematic 4–8 second video with synchronized audio, tuned for speed. Text-to-video or first/last frame.
- [Seedance 2.0 Fast](https://genkiln.com/models/seedance-2.0-fast): model ID `seedance-2.0-fast`, by ByteDance, text-to-video and image-to-video, from 77 credits ($0.077) per second of output. Seedance 2.0 Fast: 4–15 second clips at 480p or 720p with optional audio, billed per second of output.
- [Kling 3.0](https://genkiln.com/models/kling-3.0): model ID `kling-3.0`, by Kuaishou, text-to-video and image-to-video, from 91 credits ($0.091) per second of output. Kling 3.0: 3–15 second video at 720p or 1080p, with optional generated sound, billed per second.
- [Nano Banana 2](https://genkiln.com/models/nano-banana-2): model ID `nano-banana-2`, by Google, text-to-image and image-to-image, from 52 credits ($0.052) per image. Google's fast image model for generation and editing with up to 10 reference images, 1K to 4K output.
- [Nano Banana Pro](https://genkiln.com/models/nano-banana-pro): model ID `nano-banana-pro`, by Google, text-to-image and image-to-image, from 117 credits ($0.117) per image. Google's highest-quality image model: sharp text rendering, multi-image composition (up to 8 references), 1K/2K/4K.
- [GPT Image 2](https://genkiln.com/models/gpt-image-2): model ID `gpt-image-2`, by OpenAI, text-to-image and image-to-image, from 39 credits ($0.039) per image. OpenAI's GPT Image 2: prompt-faithful generation and editing with up to 16 input images, 1K to 4K.

Machine-readable list with prices and input JSON Schemas: `GET https://genkiln.com/v1/models` (no API key needed).

## Pricing

Pay as you go: buy a credit pack, then each task spends credits at the model's price. Details: https://genkiln.com/pricing

- Starter: US$10 for 10,000 credits
- Builder: US$50 for 52,500 credits
- Scale: US$200 for 220,000 credits

## Policies

- [Terms of Service](https://genkiln.com/terms-of-service)
- [Privacy Policy](https://genkiln.com/privacy-policy)
- [Refund Policy](https://genkiln.com/refund-policy)
- [Acceptable Use Policy](https://genkiln.com/acceptable-use-policy)

## Optional

- [Full documentation in one file](https://genkiln.com/llms-full.txt): every docs page, plus each model's prices and parameters.