Ship against Chinese frontier models without China phone verification, local billing setup, or VPN workflows. Use one endpoint, pay in USD with PayPal, and keep your existing OpenAI SDK integration.
OpenAI SDK compatible · Public model catalog · PayPal checkout
curl https://api.tokenoutput.cc/v1/chat/completions \
-H "Authorization: Bearer sk-your-key" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3.6-plus",
"messages": [{"role": "user", "content": "Summarize this codebase."}]
}'Use the same chat completions shape your app already speaks. Swap the base URL and start testing supported Chinese models.
Skip local account friction when you need model access fast for product experiments, side projects, and internal tooling.
Compare providers, read exact public pricing in docs, and keep using pay-as-you-go after plan limits instead of hitting a hard stop.
Both plans include strongest model access for coding workflows. The difference is monthly request volume. Overage continues with pay-as-you-go.
For solo developers and lighter coding workflows.
For frequent usage, larger codebases, and daily AI-assisted development.
Need more detail first? See full pricing details.
Keep one OpenAI-compatible integration path while testing supported DeepSeek, Qwen, and GLM public model IDs.
Plans show explicit request limits, and usage keeps flowing through pay-as-you-go instead of ending in a surprise hard stop.
Create an account, copy your API key, and move from docs to your first request in minutes.
Use docs for the current public model catalog, pricing visibility, and implementation reference.
Docs is the implementation reference for current public model IDs, pricing visibility, request examples, and integration details.
Open integration docsOne API call to the chat completions endpoint counts as one request against the current plan period.
Your usage does not stop. Overage continues with pay-as-you-go billing at standard usage rates.
Yes. Starter and Pro both include access to the strongest available model capability for coding workflows. The main difference is monthly request volume.
Yes. Both plans support up to 1M input tokens per request, with per-model limits documented in docs.
Create your account, copy your key, and use docs when you need implementation detail.
Create Free Account