What is TokenRoute

TokenRoute provides a shared API entry point for multiple model vendors. Use a platform API Key with a protocol supported by your model, and manage quota and review usage in the console.

Ways to connect

Your task	Start here
Use models in an existing client or coding tool	Integrations: configure the URL, key and model for your tool
Call models from your application	First Call: choose a protocol and language
Find model IDs and listed endpoints	Models: actual calls also depend on key permissions and routes
Configure an organization, departments and quota	Get an API Key: use the actions available to your role

One key can serve multiple protocols. Model names, message structures, tool history and advanced parameters still need to match the selected protocol.

Minimal developer friction

For OpenAI Python Chat Completions, set TOKENROUTE_API_KEY in your current terminal, then configure the client:

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["TOKENROUTE_API_KEY"],
    base_url="https://api.smartwan.com/v1",
)

See First Call for environment setup and complete requests. When migrating code, also verify model names, endpoint and parameter compatibility. Anthropic and Google SDKs use different Base URL conventions; see Authentication.

Supported models

Models displays the live platform catalog. Tests on 2026-09-22 covered text models from multiple vendors and three image-generation models; see Vendor differences.

A catalog entry, permission to call a model, and a successful request are separate conditions. Image-generation success does not establish editing, video or support for every advanced parameter. Verify your intended use case before integrating it.

Billing and cost

Models use token-based, tiered or per-call billing, with potentially different input, output and other applicable rates. Check the units in console model pricing and reconcile charges against usage details.

Client estimates and example outputs do not replace actual bills. The applicability and deduction order of paid and complimentary balances depend on the rules and billing records for your account.

Enterprise service terms

Availability targets, support hours, incident response times, retention and caching policies are governed by your signed service agreement and applicable data-processing policy. Confirm the relevant terms with the service provider before relying on these guarantees.

Next steps

Your next task	Where to go
Complete a first call	Quickstart
Obtain an API Key	Get an API Key
Configure a client tool	Integrations
Check URLs and authentication	Authentication & Base URL
Diagnose a failed call	FAQ