Is GPTokens run by a model provider?
No. GPTokens is an independent OpenAI-compatible gateway. It issues GPTokens API keys and routes requests to supported Chinese LLM providers through one account.
Pricing
GPTokens pricing and token billing
GPTokens pricing can vary by model route, account settings, provider availability, plan, and live billing configuration.
Public billing example — checked 5 September 2026
The public /api/status model_pricing record listed glm-5.3-flash at $0.12 per million input tokens and $0.40 per million output tokens. At those reference rates, 1,000,000 uncached input tokens plus 250,000 output tokens cost $0.22 before account-specific adjustments. This dated example is not a live quote: check the current route price and your account group before use.
What the estimate includes
Reference cost = uncached input tokens × input rate / 1,000,000 + cached input tokens × cache-read rate / 1,000,000 + output tokens × output rate / 1,000,000. Apply this only when the provider reports cache usage and the route uses these rates. Cache writes, other modalities, account discounts and future configuration changes require their own rules. Do not count cached input twice or add reasoning tokens again when they are already included in completion_tokens.
A reproducible request example
On 5 September 2026, a glm-5.3-flash chat-completions request asking “Reply with the single word OK.”, with temperature 0 and max_tokens 64, returned HTTP 200 and OK. Its usage reported 19 prompt tokens and 43 completion tokens, including 40 reasoning tokens, with 0 cached prompt tokens. The reference estimate is $0.00001948 using the rates above; this is not a claim about the final account debit. Token counts and outputs may vary on a later run. See the compatibility guide for the exact request and streaming check.
Live prices
Check Workspace before production because model availability and route prices can change.
Cost control
Track model name, prompt size, output length, latency, retry rate, and fallback rate to understand real cost.
Related pages
Continue from here
FAQ
Questions developers ask before buying Chinese LLM tokens
Can I use an OpenAI SDK?
Yes. Most integrations only need the GPTokens base URL, a GPTokens API key, and a supported model name from the live model list.
Who is this for?
GPTokens is built for developers and teams in the US, Europe, and other global markets that want to test Chinese LLM APIs without managing separate provider accounts, payment methods, and dashboards.