Quickstart
Sign up, create an API key, and make your first call with Python, Node.js or curl. The examples use the text model deepseek-v4-flash; the model pages show video and image requests.
Create an account
Go to fishtoken.ai/register. Choose a username and a password, enter your email address, and type in the verification code sent to that address.
Add balance
FishToken is prepaid. In the console, top up with one of the payment methods shown on the top-up page, or redeem a code. The balance does not expire.
Create an API key
Open API keys in the console and create a key. Keep it secret: do not put it in front-end code or public repositories. Store it as an environment variable:
Terminalexport FISHTOKEN_API_KEY=your-key-hereMake your first call
Use the base URL
https://fishtoken.ai/v1with any OpenAI-compatible SDK.
Python
Install the SDK: pip install openai
from openai import OpenAI
import os
client = OpenAI(
base_url="https://fishtoken.ai/v1",
api_key=os.environ["FISHTOKEN_API_KEY"],
)
resp = client.chat.completions.create(
model="deepseek-v4-flash",
messages=[
{"role": "user", "content": "Write a two-line caption for a handmade batik tote bag."}
],
)
print(resp.choices[0].message.content)Node.js
Install the SDK: npm install openai. The example uses ES modules (save it as first-call.mjs).
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://fishtoken.ai/v1",
apiKey: process.env.FISHTOKEN_API_KEY,
});
const resp = await client.chat.completions.create({
model: "deepseek-v4-flash",
messages: [
{ role: "user", content: "Write a two-line caption for a handmade batik tote bag." },
],
});
console.log(resp.choices[0].message.content);curl
curl https://fishtoken.ai/v1/chat/completions \
-H "Authorization: Bearer $FISHTOKEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v4-flash",
"messages": [
{"role": "user", "content": "Write a two-line caption for a handmade batik tote bag."}
]
}'Next steps
Generate a video
Video models use POST /v1/video/generations. Start with Wan 2.7, Kling V3, Vidu Q3 or Seedance 2.0 Fast.
Read the full API reference
Parameters, response formats and every endpoint are in the FishToken docs.
Guides
Longer walkthroughs with runnable code.
Chinese AI video APIs compared
Seedance, Kling, Wan and Vidu side by side: model IDs, modes, endpoint, and one example that switches models by changing one field.
Switch an OpenAI SDK project in two lines
Point an existing Python or Node.js project at FishToken with base_url and api_key, or with two environment variables.
Seedance 2.0 Fast in Python
A complete script: submit a generation, poll the task, save the result URL. With prompt tips.
If something goes wrong
- Authentication error
- Check that the request sends
Authorization: Bearerfollowed by your FishToken key, and that the base URL ishttps://fishtoken.ai/v1. - Model not found
- Use the model ID exactly as listed on the models page, for example
deepseek-v4-flash. - Balance or quota error
- Add balance in the console, then retry.
- Too many requests
- Standard accounts can send 60 requests per minute and accounts that have topped up can send 120. Slow down and retry with a short back-off.
- Still stuck
- Email support@fishtoken.ai with the time of the request and the error message. Do not send your API key.
Get your API key
Create an account, add balance, and make your first call from the quickstart.