Every model.
One key.
EasyToken puts Veo, Seedance, Kling, GPT Image, Suno, Claude and more behind a single API key and a single bill. You get the same models at lower rates, and a failed job costs you nothing.
Same models. Smaller bill.
Pay per output, from prepaid credits. No seats, no minimums. Top up when you need to, and see every charge per task in the dashboard.
| Model | Billed per | Official | EasyToken | Saving |
|---|
Launch rates in USD. Live per-model pricing is always shown in your dashboard before you run a task. Browse all models →
From key to first video in three calls.
Media jobs run asynchronously, so a long Veo render never ties up your server. Chat models answer in the same request.
Create a key
Sign up, claim your free credits, and copy one key that works for every model.
Send a task
POST to /v1/tasks with a model and its input. Switching models means changing one string.
Get the result
We call your callback_url when the file is ready, or you poll /v1/tasks/{id}.
Built for teams that ship AI features.
You shouldn't need five vendor accounts, five invoices and five SDKs to put video, voice and chat in one product.
Automatic fallback
If an upstream provider stalls or rate-limits, we retry on a healthy route so your users don't see an error.
Pay only for results
A task that fails or gets filtered is refunded automatically. You are charged for files you receive.
One bill
One credit balance and one invoice across every model, with cost broken down per task, per key and per day.
Spend caps per key
Give each app, client or teammate its own key with a daily limit, and revoke it in one click.
Playground
Test prompts against any model in the browser, then copy the exact request as curl, Python or JavaScript.
Your files stay yours
Outputs are stored for 14 days for download, then deleted. Prompts are never used for training.
Already using the OpenAI SDK? Change two lines.
Our chat endpoint speaks the OpenAI format, so existing code, LangChain apps and coding agents work after you swap the base URL and key.
- Python, Node, Go, or plain curl
- Streaming, tool calls and JSON mode
- Same key for video and image tasks
from openai import OpenAI client = OpenAI( base_url="https://api.easytoken.si/v1", api_key="et_live_...", ) reply = client.chat.completions.create( model="claude-opus-5-5", messages=[{"role": "user", "content": "Write a 15s ad script"}], ) print(reply.choices[0].message.content)
Questions before you switch.
Are these the official models?
Yes. EasyToken routes your request to the model provider's own endpoint or an authorised host. The output is the same model you would get directly.
How can you be cheaper than the official API?
We buy capacity in volume across many customers and pass most of the discount on. Prices are listed per model, and you see the cost before each task runs.
What happens when a task fails?
Failed, timed-out and safety-filtered tasks are refunded to your balance automatically. You don't need to file a ticket.
Do credits expire?
No. Prepaid credits stay on your account until you use them.
Are there rate limits?
Every account starts with generous concurrency per model. If you need more for a launch, contact us and we raise it the same day.
Can I get an invoice for my company?
Yes. Add your company name and VAT number in billing settings and every top-up generates a proper invoice.
Start with free credits.
Create a key in under a minute and run your first video, image or chat request today.