Provider routing
Route each request through approved providers and models, with ordered fallback, health probes, circuit breaker decisions, and request-level records.
AI token routing infrastructure
One API layer for models, credits, and agent traffic.
Tokenok gives global developers and AI agents an OpenAI-compatible gateway for smart model routing, token credits, usage metering, checkout, invoices, and multilingual customer operations.
Base URL
https://api.tokenok.net/v1
Drop it into OpenAI-compatible SDKs and keep your model policy in Tokenok.
Paid revenue
148,200
CREDITS
Failure rate
0.42%
last 24h
openai / tokenok-fast
healthy
anthropic / fallback
degraded
Checkout and invoices
Live
signed webhook settlement
Base URL
api.tokenok.net
Unified API and settlement layer
The control plane for selling and operating AI model access globally.
Token credits and invoices
Track token cost, deduct prepaid balances, reconcile checkout sessions, verify signed payment webhooks, and issue invoice PDFs.
Customer operations
Let customers manage API keys, recharge orders, billing profiles, request logs, statements, invoices, and onboarding readiness.
Global console
Operate admin and customer workspaces in Chinese, English, Japanese, and Korean from the same console.
Model coverage
Global flagship models and China-native models in one routing catalog.
Start with curated presets, then configure provider Base URLs, encrypted API keys, prices, aliases, and routing priorities in the admin console.
OpenAI, Anthropic, Google Gemini, xAI, and Mistral presets cover premium, coding, fast, and low-cost routing groups.
Alibaba Qwen, DeepSeek, Moonshot Kimi, Z.ai GLM, and MiniMax presets are ready for OpenAI-compatible Base URL and key setup.
Commercial readiness
One account, hundreds of models, one invoice.
Launch workflow
From demo access to paid usage without custom spreadsheets.
Configure providers
Add Base URLs, encrypted API keys, model prices, routing aliases, and provider health probes from the admin console.
Invite customers
Create customer accounts, assign provider and model policy, issue API keys, and start with demo access or paid credits.
Operate revenue
Track request logs, balances, failed calls, recharge orders, payment webhooks, invoices, and audit logs in one workspace.
Who it serves
Built for teams turning model access into a managed token service.
AI platform operators
Centralize provider routing, account policy, balance controls, and operational audit trails before inviting customers.
AI agencies and integrators
Offer clients a branded gateway with prepaid usage, invoices, and request logs instead of manual spreadsheet reconciliation.
Internal AI enablement teams
Give business units controlled model access, visible usage, and safer key lifecycle management across regions.
Commercial FAQ
The details buyers check before moving traffic.
Can existing OpenAI clients migrate?
Yes. Set the Base URL to `https://api.tokenok.net/v1`, use a Tokenok API key, and request a model alias such as `tokenok-fast`.
How do customers pay?
Launch accounts can use prepaid CREDITS, manual settlement, or enabled online checkout providers with signed webhook reconciliation.
What does the operator control?
Admins control provider keys, model availability, routing policy, customer balances, API key lifecycle, announcements, and audit trails.