Token Wallet API

Base URL

https://token-wallet.ai/api/v1

Every endpoint below is relative to this base.

Authentication

Create a key on the API Keys page. Keys are shown once at creation — store yours in a secret manager. Send it as a bearer token in the Authorization header.

curl https://token-wallet.ai/api/v1/chat/completions \
  -H "Authorization: Bearer tw_xxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4-flash",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

Chat completions

POST /chat/completions is fully OpenAI-compatible: just model and messages. Calls draw down your subscription quota; model IDs come from GET /models.

{
  "model": "deepseek/deepseek-v4-flash",
  "messages": [
    { "role": "system", "content": "You are concise." },
    { "role": "user", "content": "Summarise Singapore's AI policy in 3 bullets." }
  ],
  "temperature": 0.5,
  "max_tokens": 512,
  "stream": false
}

The response is a standard OpenAI chat.completion object:

{
  "id": "chatcmpl-…",
  "object": "chat.completion",
  "created": 1783200000,
  "model": "deepseek/deepseek-v4-flash",
  "choices": [
    {
      "index": 0,
      "message": { "role": "assistant", "content": "…" },
      "finish_reason": "stop"
    }
  ],
  "usage": { "prompt_tokens": 24, "completion_tokens": 86, "total_tokens": 110 }
}

Set "stream": true for token-by-token SSE output — standard chat.completion.chunk frames ending with data: [DONE]; the final frame carries usage.

List models

GET /models returns the active models and their per-1M token prices.

curl https://token-wallet.ai/api/v1/models -H "Authorization: Bearer tw_xxxxxxxxxxxx"

{
  "object": "list",
  "data": [
    {
      "id": "deepseek/deepseek-v4-flash",
      "object": "model",
      "owned_by": "ppio",
      "input_price_per_1m_tokens": 0.19,
      "output_price_per_1m_tokens": 0.37
    }
  ]
}

Code examples

Python (official OpenAI SDK):

from openai import OpenAI

client = OpenAI(base_url="https://token-wallet.ai/api/v1", api_key="tw_xxxxxxxxxxxx")

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

Node (official OpenAI SDK):

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://token-wallet.ai/api/v1",
  apiKey: "tw_xxxxxxxxxxxx",
});

const r = await client.chat.completions.create({
  model: "deepseek/deepseek-v4-flash",
  messages: [{ role: "user", content: "Hello" }],
});
console.log(r.choices[0].message.content);

GUI clients (Cherry Studio, LobeChat, …): add an “OpenAI-compatible” provider with the base URL above and your tw_ key.

Errors

Errors return { "error": { "message", "type" } } with the matching HTTP status.

• 400Invalid input — model or provider unknown, or schema mismatch.
• 401Missing or invalid API key.
• 402No active pack for this provider/model, or the pack ran out mid-call. Buy a pack at /packs.
• 403Account suspended, or this model isn't allowed for this key.
• 429Rate limit exceeded. Slow down and retry shortly.
• 502Upstream provider error. The call was not charged.

Rate limits

Default: a burst of 10 requests, refilling at ~0.5 requests/second per key. Need more? Get in touch.

Security

• Keys are hashed before storage — we never see the plaintext after creation.
• Pack deductions run inside DB transactions; failed provider calls are never charged.
• Upstream provider keys live only on the server. Never expose them to the browser.
• Revoke a key from the API Keys page the moment you suspect it has leaked.