# Wokey AI Wokey AI provides a single API key for models from providers including Anthropic, OpenAI, Moonshot AI and xAI, with compatible endpoints and locally verifiable response proofs. - Findings24: - 网站: - 价格: Freemium - 分类: AI 网关与路由, 大语言模型 API ## 什么是 Wokey AI？ Wokey AI is an API relay for accessing multiple language-model providers through one API key. It supports OpenAI-compatible Chat Completions and Responses endpoints, as well as Anthropic-compatible Messages, so developers can route supported applications through Wokey by changing the base URL and using a Wokey key. The service states that requests go to official provider APIs and supplies cryptographic response proofs that can be checked locally, including an attestation chain to AWS Nitro. ## Wokey AI 能做什么？ ### One key across model providers Use a single Wokey API key to access listed models from providers such as Anthropic, OpenAI, Moonshot AI and xAI. ### Compatible API protocols Supports OpenAI Chat Completions and Responses, plus Anthropic Messages. Existing compatible clients can be configured with a different base URL. ### Offline response verification Responses carry cryptographic proofs that can be checked locally with Wokey's open-source verifier, without relying on the relay operator's word. ### Attestation and signature checks The verification workflow checks the code fingerprint against audited source, validates the response signature against the attested key, and traces the attestation to the AWS Nitro root. ### Usage-based billing and model rates Pricing is listed per million input and output tokens, with cache-read rates shown for supported models. Usage is itemized; some models have scheduled peak and off-peak rates. ### Automatic failover Wokey describes pooled model supply and smart routing that can fail over automatically. ## 使用场景 ### Switch an existing API client Developers using compatible OpenAI or Anthropic clients can point the base URL at Wokey and authenticate with a Wokey API key rather than rewriting their request workflow. ### Verify model responses independently Teams that need to inspect the provenance and integrity of API output can run the verifier locally and offline against the response proof. ### Compare model costs Application teams choosing among supported models can consult per-token input, output and cache-read rates alongside the listed official rates. ### Configure coding assistants Users can connect Claude Code or Codex by setting the client base URL to Wokey and supplying their Wokey key, following the product's setup guides. ## 常见问题 ### How do I connect an existing client? Create a Wokey API key, set the client's base URL to Wokey, and use the key for authentication. Compatible OpenAI Chat Completions, Responses, and Anthropic Messages workflows are supported. ### How can I verify a response? Use Wokey's open-source relay verifier locally. It works offline and checks the response signature and hardware-backed attestation chain. ### How is API usage priced? The pricing page lists rates in USD per one million tokens, including input, output and, where listed, cache-read rates. Billing is described as per-token with itemized usage; some models have scheduled peak and off-peak rates. ### Can I use Wokey with Claude Code or Codex? Yes. The published setup approach is to point the client's base URL at Wokey and use a Wokey API key; Wokey provides guides for both clients. ## 快速信息 - **Product type**: Multi-provider language-model API relay - **API protocols**: OpenAI Chat Completions and Responses; Anthropic Messages - **Verification**: Local, offline verification of cryptographic response proofs - **Billing basis**: Per-token usage; rates are listed per 1 million tokens - **Top-up method**: Multi-chain USDT top-ups - **Website**: wokey.ai