> ## Documentation Index
> Fetch the complete documentation index at: https://yomo.run/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat Completions API

> LLM Bridge provides an OpenAI API-compatible API server with support for multiple providers.

## API

LLM Bridge provides [POST /chat/completions](/api-ref/endpoint/post) API endpoint.

## LLM Bridge Configuration

```yaml yomo.yaml theme={null}
name: llm-bridge
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: localhost:9000
      provider: openai

    providers:
      openai:
        api_key: sk-proj-xxxxx
        model: gpt-4.1
```

## Configuration Fields

### API Server Configuration (bridge.ai.server)

* `addr`: The address of the API server.
* `provider`: The provider of the API server. Supported providers include `openai`, `ollama`, `vllm`, `gemini`, `anthropic`, `vertexai` and etc.

### Providers Configuration (bridge.ai.providers)

#### OpenAI Provider

* `api_key`: The API key for OpenAI. You can grab it from the [OpenAI Dashboard](https://platform.openai.com/account/api-keys).
* `model`: The model to use. Supported all chat completions models like `gpt-4.1`, `gpt-o4` and etc.

```yaml theme={null}
name: ai-zipper
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: localhost:9000
      provider: openai
      
    providers:
      openai:
        api_key: sk-xxxxxxxxxxxxxxxxx
        model: chatgpt-4o-latest
```

#### Ollama Provider

* `api_endpoint`: The endpoint of the Ollama API server. Default is `http://localhost:11434`.

```yaml theme={null}
name: ai-zipper
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: localhost:9000
      provider: ollama
      
    providers:
      ollama:
        api_endpoint: http://localhost:11434
```

#### Google Gemini Provider

* `api_key`: The API key for Google Gemini. You can grab it from the [Google Cloud Console](https://console.cloud.google.com/apis/credentials).

```yaml theme={null}
name: ai-zipper
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: localhost:9000
      provider: gemini
      
    providers:
      gemini:
        api_key: xxxxxxxxxxxxxxxxx
```

#### vllm Provider

* `api_endpoint`: The endpoint of the vllm API server. Default is `http://localhost:9999/v1`.
* `api_key`: The API key for vllm.
* `model`: The model to use. Supported all chat completions models like `deepseek-ai/DeepSeek-R1`, `llama3` and etc.

```yaml theme={null}
name: ai-zipper
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: localhost:9000
      provider: vllm
      
    providers:
      vllm:
        api_endpoint: http://10.10.10.10:9999/v1
        api_key: xxxx
        model: deepseek-ai/DeepSeek-R1
```

#### Anthropic Provider

* `api_key`: The API key for Anthropic. You can grab it from the [Anthropic Dashboard](https://console.anthropic.com/api-keys).
* `model`: The model to use. Supported all chat completions models like `claude-3.7-sonnet`, `claude-3` and etc.

```yaml theme={null}
name: ai-zipper
host: 0.0.0.0
port: 9000

bridge:
  ai:
    server:
      addr: localhost:9000
      provider: anthropic
      
    providers:
      anthropic:
        api_key: sk-ant-api03-xxxxx
        model: claude-3.7-sonnet
```