OpenAI-compatible API
Core Facts
Timeline (last 90 days)
The app's communication architecture is split into two channels: the Android client interacts with the Agent through the local OpenAI-compatible API provided by Hermes; a separate control API manages the lifecycle of the virtual machine and the Agent.
Spring AI and LangChain4j are both mature LLM integration options on the JVM, and both support OpenAI-compatible APIs, covering a large number of hosted models and local model services (such as Ollama and vLLM).
The Dynamo frontend exposes an OpenAI-compatible API, clients call it using familiar routes, and the frontend routes requests to backend Workers.
Locally Uncensored exposes an OpenAI-compatible local API on port 8129, which by default only listens on localhost and requires token authentication.
LLST interfaces via the token mode of the OpenAI-compatible API, supporting multiple inference frameworks.
All Facts (6)
OpenAI API格式已成为大模型领域事实上的行业标准接口规范，其核心是/v1/chat/completions端点
80%UnverifiedThe app's communication architecture is split into two channels: the Android client interacts with the Agent through the local OpenAI-compatible API provided by Hermes; a separate control API manages the lifecycle of the virtual machine and the Agent.
50%UnverifiedSpring AI and LangChain4j are both mature LLM integration options on the JVM, and both support OpenAI-compatible APIs, covering a large number of hosted models and local model services (such as Ollama and vLLM).
50%UnverifiedThe Dynamo frontend exposes an OpenAI-compatible API, clients call it using familiar routes, and the frontend routes requests to backend Workers.
50%UnverifiedLocally Uncensored exposes an OpenAI-compatible local API on port 8129, which by default only listens on localhost and requires token authentication.
50%UnverifiedLLST interfaces via the token mode of the OpenAI-compatible API, supporting multiple inference frameworks.
50%