Frequently asked questions
What is TokenMilo?
TokenMilo is an API service that gives you access to 21 Claude and GPT models — including Claude Fable, Opus, Sonnet, Haiku and GPT-6 — through one API key, at prices well below official rates.
Are these the full, official models?
Yes. Requests are served by the real Claude and GPT models. We do not swap in smaller, quantized or downgraded models.
How much cheaper is it than the official API?
Claude models are at least 77% cheaper than official Anthropic pricing, and GPT models at least 89% cheaper than official OpenAI pricing. See the pricing page for exact per-model prices.
How do I pay?
You can top up with Alipay, WeChat Pay. No overseas credit card is needed. Billing is pay-as-you-go with no subscription.
Which APIs are supported?
The Anthropic Messages API (https://api.tokenmilo.com/v1/messages) and the OpenAI-compatible Chat Completions and Responses APIs (https://api.tokenmilo.com/v1). Existing SDKs work by changing the base URL and API key.
Can I use it with Claude Code?
Yes. Set ANTHROPIC_BASE_URL to https://api.tokenmilo.com and ANTHROPIC_AUTH_TOKEN to your TokenMilo key, then start Claude Code. See the quickstart for details.
Does it support prompt caching?
Yes. Cache reads and cache writes are billed at the cache prices listed on the pricing page, which are far cheaper than regular input tokens for repeated context.