Documentation

Everything you need to call Chinese frontier models through HeyToken.

Quick start

1. Get your API key

Register an account, top up your wallet, then create a key in the console. Keys look like sk-....

2. Point your SDK at HeyToken

HeyToken is wire-compatible with the OpenAI chat completions API:

from openai import OpenAI

client = OpenAI(
    base_url="https://api.heytoken.io/v1",
    api_key="sk-your-heytoken-key",
)

resp = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)

3. Call any model

Set the model parameter to any model ID from the models page, e.g. deepseek-v4-flash, glm-5.3, kimi-k2.7-code.

Authentication

Send your key as a Bearer token in the Authorization header. Keep keys server-side; never ship them in client code.

Authorization: Bearer sk-your-heytoken-key

Error codes

Errors follow the OpenAI error format. Common cases:

  • 401Invalid or missing API key.
  • 402Insufficient wallet balance. Top up to continue.
  • 429Rate limited or upstream busy. Retry with backoff.
  • 500Upstream provider error. Retried automatically across channels when possible.

Available models

The full, always-up-to-date list with pricing lives on the models page.

Browse models

More questions? Check the FAQ on the home page.