SDK quickstart

One API. Your language.

Install TokenSpend for Python or TypeScript. Every request lands in the same receipt history.

Get a rail key API reference

SDK quickstart

One API. Pick your language.

TokenSpend clients use one rail key, model catalog, credit balance, and receipt history.

Model

Search by maker, model, or ID.

Use the TokenSpend Python client

pip install tokenspend

pip install tokenspend

from tokenspend import TokenSpend

client = TokenSpend()
response = client.chat.completions.create(
    model="zai/glm-5.2",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Your first successful request will appear in Router activity.

Key: $TOKENSPEND_API_KEY

API

api.tokenspend.dev

Authentication

One tsr_ rail key

Proof

Request ID + receipt

Safe retries

One job, one charge.

Reuse a key only for the same request. Duplicate attempts return the original request ID.

Python

response = client.chat.completions.create(
    model="zai/glm-5.2",
    messages=messages,
    idempotency_key="deploy-42-generation",
)
print(response.request_id)

TypeScript

import { getResponseMetadata } from "@tokenspend/sdk";

const response = await client.chat.completions.create(
  { model: "zai/glm-5.2", messages },
  { idempotencyKey: "deploy-42-generation" }
);
console.log(getResponseMetadata(response)?.requestId);